GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted
Musings of an Energy Nerd

All About Attic Venting

We vent attics for four reasons, and all four goals can be better achieved by adopting measures other than attic venting

Unless you are building a sealed, conditioned attic, the building code requires attics to be ventilated. However, air isn't always smart enough to follow the “smart arrows” shown in ventilation illustrations like this one.
Image Credit: Fine Homebuilding

Most homeowners and builders believe that attics should be vented. If you walk down to your local lumberyard and lean on the counter, the employees and nearby customers will offer a variety of opinions about why attics need to be vented. Unfortunately, it’s highly unlikely that the statements you hear will be true.

Here are the four most common reasons people suggest to explain the practice of venting attics:

  • To reduce the chance of moisture build-up in the attic or condensation on the underside of the roof sheathing.
  • To make roofing shingles last longer.
  • To lower cooling bills during the summer.
  • To reduce the chance of ice dams.

Although attic ventilation is sometimes able to contribute in a very small way to addressing the problems on this list, there are much better solutions to all four problems than ventilation.

What does the code require?

If you plan to install insulation on your attic floor, then most building codes require that the attic be vented. (For example, see Section R806.1 of the 2015 International Residential Code.)

The standard code formula requires 1 square foot of net free ventilation area for every 300 square feet of attic floor area, assuming that half of the ventilation openings are located in the lower half of the attic (generally at the soffit) and half near or at the ridge. If a roof has only soffit vents and no ridge vents, most codes require 1 square foot of net free ventilation area for every 150 square feet of attic floor area.

Manufacturers of soffit vents and ridge vents usually specify the net free vent area of their products on product packaging or in specifications available online. (Researchers have shown that the net free vent areas reported by manufacturers are exaggerated, but that is a topic for another article.)

If you…

GBA Prime

This article is only available to GBA Prime Members

Sign up for a free trial and get instant access to this article as well as GBA’s complete library of premium articles and construction details.

Start Free Trial


  1. user-757117 | | #1

    A vented attic is like a sloped rainscreen...
    Seems like the same rationale that applies to recommending a rain screen gap should also apply to recommending a vented attic.

  2. GBA Editor
    Martin Holladay | | #2

    Response to Lucas Durand
    There are similarities between vented cathedral ceilings and walls with a rainscreen gap -- and there are differences. Attics, of course, are a whole 'nuther kettle of fish.

    1. Roofs can have snow on them; walls generally don't.

    2. Air leakage into roof assemblies is much more likely than air leakage into wall assemblies, due to the stack effect. That means that cathedral ceilings face a bigger moisture load due to air leaks than walls do.

    The analogy is useful, but we shouldn't let the similarities between these assemblies blind us to the fact that there are important differences. We need to look at data, not analogies.

    That said, examples of rotten OSB roof sheathing in unvented cathedral ceiling assemblies certainly show that a ventilated air gap can make a big difference -- in some cases, the difference between a system that works and failure.

  3. user-1140531 | | #3

    Regarding this statement by Jeff Gordon:

    “You cannot cool the upper part of a cathedral ceiling roof with venting. “

    The statement is made to sound unconditional, and yet he qualifies it with the following explanation:

    “If a cathedral ceiling cavity is vented top and bottom, and faces south, then air moves through the cavity driven by buoyancy. Air enters at the soffit. As it moves up the vent slot it becomes heated and exits the ridge as heated air. If a cavity has a slot [that is, an air gap between the top of the insulation and the underside of the roof sheathing] but no vents, the air in the slot will have a certain temperature. The air in the vented slot will reach that same temperature after, say, 10 feet. The remaining higher part of the slot will receive no cooling effect from the moving air.”

    His explanation begins with a proper description of a vent system that includes an air passage plus inlet and outlet vents. But, then he says that if the vent system has an air passage, but no vents, it will not work to cool the roof, thus supporting his original premise that you cannot cool the upper part of a cathedral ceiling roof with venting.

    It seems to me that the explanation is saying you cannot cool the roof with venting if no venting system exists. A venting system requires an air passage plus and inlet and outlet. Without the inlet and outlet, no venting system exists.

    1. HippieBallDoo | | #70

      Actually, his explanation is quite specific, and while accurate, requires carefully keeping track of the details. He is switching context on you between the unvented and vented channels. You've simply lost track of his switches.

      Gordon states that in an unvented roof, the air is stagnant. In a vented roof, air does move, but quickly reaches the SAME temperature as the unvented roof due to thermal transfer from the sun on the shingles. Therefore, he concluded "the slot will receive no cooling effect from the moving air," as in vented or unvented the channel temperature at the top is the same. Yes, the channel at the bottom in the vented roof does cool, but that cool air is quickly heated to the same temperature as the unheated air.

      Therefore, the correct roof philosophy is not to change the temperature, but to control the moisture. Vented or unvented, keep the moisture out.

  4. GBA Editor
    Martin Holladay | | #4

    Response to Ron Keagle
    Jeff Gordon may have expressed the idea inelegantly or clumsily, but he was reporting actual temperature measurements made in vented cathedral roof assemblies. The data come from a study performed by Bill Rose; here is a link to his paper: Measured Summer Values of Sheathing and Shingle Temperatures for Residential Attics and Cathedral Ceilings.

    I reported on this study in the July 2002 issue of Energy Design Update:

    "Rose has developed his own statistical method for comparing temperatures from roof to roof. The data were represented in a form of correlational analysis. This was done using linear regression of individual cases against the base case. The slope of the regression line was used to permit the temperature of the comparison case to be expressed as “hotter than” or “colder than” the base case as a percent. The statistical method behind these comparisons is fairly complex. The method’s most important feature is its usefulness in ranking factors in order of their effect on shingle temperature. ...

    "Rose also found that shingles above a vented cathedral ceiling are cooler at the eaves and hotter at the ridge than shingles above a vented attic. This is due to the strong temperature gradient, especially on the south side of the roof, which exists in the ventilation channel above a cathedral ceiling. “It becomes apparent that venting can cool the lower section of a vented cathedral ceiling quite effectively, but the cooling effect is greatly reduced for the upper part of the cavity,” writes Rose.

    "There are many good reasons to vent a roof — to maintain a shingle warranty, to remove moisture from the roof assembly, and to comply with code requirements. However, when it comes to lowering shingle temperatures, ventilation is far less important than roofing color."

  5. GBA Editor
    Martin Holladay | | #5

    Second response to Ron Keagle
    Because Jeff Gordon's explanation of the temperature gradients in cathedral ceiling ventilation gaps was confusing, I have rewritten that section of my article. Thanks for your comments.

  6. user-1140531 | | #6

    Cooling Distribution on Cathedral Ceiling Roof

    Thanks for that clarification of Jeff Gordon’s statement. I looked at the paper by Bill Rose, and his finding of the strong temperature gradient in the vent air passage of cathedral ceilings. I will study it further, but did not yet notice if the area of the air passages had been stated. But, I suspect that the strong temperature gradient is the result of the area of the air passages slowing the upward airflow due to constricting the natural convection flow.

    In other words, the restriction of the passage slows down the upward airflow, so it remains in the passage for a longer time. The longer it remains in the passage, the more it is heated. The warmer the air becomes, the slower the heat transfers from the shingles to the air.

    Thus, the lower air is heated rapidly due to the temperature difference, and the heat transfer slows as the rising air gets hotter. The slower the heat transfer from the shingles to the ventilating air, the hotter the shingles become from the solar gain.

    If you increase the cross sectional area of the air passage, a greater quantity of cooling air will enter and move through the passage, and thus have added capacity to absorb more heat transfer from the shingles higher in the air passage; therefore, reducing the temperature gradient of the shingles between the lower and higher regions of the roof shingles.

    So, while there will always be a temperature gradient between the lowest and highest part of the roof, that gradient can be minimized by reducing airflow constriction in order to maximize the airflow rate. I see no reason why the gradient cannot be made so small that considerable cooling of the upper part of the roof can be achieved.

    So, I disagree with Jeff Gordon when he says, “You cannot cool the upper part of a cathedral ceiling roof with venting.”

  7. GBA Editor
    Martin Holladay | | #7

    Third response to Ron Keagle
    Bill Rose explains the gap in his experimental cathedral ceilings this way: "Nominal 10 in. R-30 fiberglass batt insulation with kraft facing is fastened at the bottom of the 2×12 framing cavity. The resulting airspace between the top of the batt and the underside of the sheathing was not controlled to a uniform dimension. There were no air chutes in the cathedral ceilings."

    Since 2x12s measure 11.25 inches, the ventilation gaps were roughly 1.25 inch high.

    If you are hoping that ventilation channels will significantly change the temperature of your asphalt shingles, you are likely to be disappointed. All research data on this topic are consistent: cathedral ceiling vents only lower asphalt shingle temperatures very slightly. The color of the shingles matters much more than whether or not there is a ventilation gap under the sheathing. Even when shingles are installed over a vented attic -- one in which the vent channel is basically between 6 inches and 8 feet high -- venting the attic is less important than shingle temperature, geographical location, and the orientation of the roof slope.

  8. Cjb80 | | #8

    Vented Cathedral Ceiling
    (I apologize in advance if this is too far off topic)

    I have been designing a home with a vented cathedral ceiling for some time now, and I am all but complete at this point (I am working on the final drawings). I recently revised the overhangs on the home so that they are separate construction that are fastened until the side of the home so that the sheathing can run continuously up the side of the exterior. Then it occurred to me that the house wrap could be installed underneath the overhang structure. Furthermore, the house wrap could then be extended over onto the roof into the ventilated cavity.

    The be clear, current design has 2x4 "furring strips" installed on the flats on the first layer of roof sheathing. Then a second layer will be installed on top of the 2x4s creating a 1.5" gap.

    Would it be reasonable to use house wrap underneath the 2x4 "furring strips", just above the bottom roof sheathing? The primary purpose of this would be to eliminate any wind-washing issues.



  9. user-757117 | | #9

    Response to Martin.
    Your points about the differences between assemblies are understood...

    It seems like the benefits of ventilated exterior wall cladding are well enough established that they are seldom seriously questioned...
    Is it fair to say then that if the benefits of ventilated exterior cladding are (for all intents and purposes) beyond debate, that roof assemblies should also (for all intents and purposes) benefit from ventilation in the same ways?

    I agree that data and research provide the best guide for decision making in this regard.
    However, in terms of expressing what the data and research have to say, I think "rules of thumb" are a more effective means of communication than direct references to the actual science (ie, a litany of evidence-based reasoning is seldom very useful outside a "nerdy" niche audience if it can't be reduced for consumption by a wider "less nerdy" audience).
    The trick, I think, is to develop new "rules of thumb" that accurately reflect the state of the underlying science and then work them into the existing culture over time until they are essentially matters of faith (which is all a lot of the present "rules of thumb" are).

    To be clear, I'm just musing here and wouldn't want you to take my comments as being critical of the detail within your blog post.

  10. GBA Editor
    Martin Holladay | | #10

    Response to Chris Barnes
    If I understand you correctly, you are building your house without roof overhangs, in order to simplify the air-sealing details at the exterior air barrier. This is a good approach.

    Most builders who follow this approach use tape to create the air barrier. The easiest way to do this is to use Zip System sheathing and Zip System tape. If you follow this approach, you won't have any air leakage at the sheathing seams. Of course, you also need to seal the seam at the top of the wall, where the wall meets the roof sheathing. As long as you follow this approach, you don't need to fold your housewrap onto your roof.

    For more information on this approach, see Airtight Wall and Roof Sheathing.

  11. GBA Editor
    Martin Holladay | | #11

    Response to Lucas Durand
    Basically, I agree with you. But there are a few caveats.

    Rainscreen gaps on walls have many benefits and few downsides (other than cost and the need to develop new flashing details). If you add a ventilated rainscreen gap on the outside of your wall sheathing, the gap won't cause any problems.

    The same cannot be said for a cathedral ceiling. Increasing the vent openings (soffit vents and ridge vents) that connect to a cathedral ceiling has definitely been known to tip a functioning roof assembly into failure. That's because a new ridge vent can pull more interior air through ceiling leaks, overloading the assembly with interior moisture.

    All of these issues are connected. A ventilation gap in a cathedral ceiling will only work if the ceiling is airtight. Without the airtight ceiling, the ventilation makes things worse.

    I'm sure that you understand these issues, but many builders don't. When we give advice to builders, we need to encourage whole-systems thinking, so that builders know that the thigh bone is connected to the knee bone.

    When it comes to attic ventilation, I tend to agree with Bill Rose. My attitude is, "Include it if you want to, or omit it -- but don't sweat the details. It doesn't matter very much."

    When it comes to cathedral ceilings, a ventilation channel is much more important than it is for an attic. But cathedral ceilings are extremely risky, and need to be detailed with care. If you insist on having cathedral ceilings in your house, you had better get the details right. And I think that a ventilation gap is a good idea, because it can make the difference between a roof assembly that works and one that doesn't.

    None of this advice makes any sense, however, unless you have an airtight ceiling. So that's the message we need to keep repeating: Get your ceiling airtight.

  12. user-757117 | | #12

    Response to Martin.
    Your conclusion makes sense.
    Thanks for helping me find perspective...

    Next time I'm at the lumber yard talking with the guys and the subject of attic ventilation comes up (as has sometimes happened), I'll know what direction to steer the conversation in ;-)

  13. Expert Member
    MALCOLM TAYLOR | | #13

    …as Lucas smiles to himself knowing his house has a service core on the ceiling with an airtight layer of OSB and will do just fine.

  14. pdunnigan | | #14

    Insulating a Knee Wall
    I have a 50's Cape Cod here in the North East. Knee Walls upstairs with a Cathedral ceiling running about 60% the length of the house with an un-insulated attic space for the rest. I know I have batt insulation behind the cathedral ceiling, but no venting for air passage to the ridge vent. I have batt insulation behind the knee wall with batt insulation carrying down the rafters to the soffit vents, but know insulation on the floor behind the knee walls. I can see my top plate and want to blow in insulation down my walls and on the floor behind the knee walls and in the attic space above my second floor. I'm not sure how to block out the air from the outside. I know I have to block out the end of the floor joists to keep air from going under my floor and ceiling. but what am I doing where the roof meets the top plate?

  15. GBA Editor
    Martin Holladay | | #15

    Response to Patrick Dunnigan
    Above the top plate of your exterior walls, you need to install a ventilation baffle on the underside of your roof sheathing between each pair of rafters or trusses. You can make your own baffle out of thin plywood or rigid foam, spaced off the sheathing 1 inch or 2 inches, or you can buy a commercial product like AccuVent.

    Then you need to install an insulation dam between the top plate of the wall and the underside of your ventilation baffle. The easiest material to use for an insulation dam is rigid foam. Once the insulation dam is installed, seal the perimeter of each dam with canned spray foam so that the installation is airtight.

    Of course, this approach assumes that you have access to the small attic behind the kneewall, and that there is enough room back there for you to work. If the access is too tight from the interior, you may have to remove the soffits on the exterior of your house, and work from the exterior.

    Here is a link to a video by Mike Guertin showing one way to do the work you are asking about: How to Ventilate Rafter Bays When Adding Insulation.

    I will post some photos and drawings below to illustrate what an insulation dam looks like.


  16. pdunnigan | | #16

    Response to Martin Holladay, GBA Advisor
    Thanks for the response, I attached a couple of pictures of my knee wall, it is big enough to get behind. I have insulation in the knee wall and going up to the peak(with a ridge vent), but no existing rafter baffles in the cathedral ceiling part of the roof(about 60% of the roof, the other 40% I can get to the attic above). I would have to rip out all the drywall to get those in the ceiling. can I do the knee wall without having to do that? I have a 2x4 on top of my rafters behind the knee wall, somehow I would need that rafter baffle to go from between the rafters all the way to the roof sheathing? Can I fill behind the Knee wall with cellulose once the baffles are placed?

  17. GBA Editor
    Martin Holladay | | #17

    Response to Patrick Dunnigan
    Your photos show a typical situation in this type of Cape home. However, just because it is typical, doesn't mean it isn't a disaster.

    The photo shows that you have the worst-performing type of insulation available: fiberglass batts.

    The photo shows that your kneewall lacks an attic-side air barrier (although, since the kneewall is entirely inside your home's thermal barrier -- kind of -- this doesn't really matter).

    The photo shows that there is no air barrier on the underside of your roof insulation.

    The photo shows that there are gaps that provide plenty of opportunities for air to flow from your soffit vents into the conditioned space of your home.

    Your description tells us that you have no ventilation baffles, so there is no guarantee that there is an air gap between the top of you roof insulation and the underside of your roof sheathing.

    To fix all these problems is a lot of work -- and yes, to fix them the right way will require you to take down the drywall on your sloping ceiling.

    You can either do it right, or you can do it halfway and cross your fingers. Maybe you'll have a moisture problem affecting your roof sheathing, and maybe you won't. Maybe you'll have ice dams, and maybe you won't. One thing's for sure, though: if you don't address these problems, your energy bills will be higher than they should be.

    More information here: Two ways to insulate attic kneewalls.

  18. doug_horgan | | #18

    A couple of observations from the DC area
    Martin, as usual your article is thorough and comprehensive. There are certainly a number of misunderstandings “out there” and you address many of them.

    I’ve had the pleasure of working on hundreds of homes in the DC area and would like to mention a few things I’ve seen over the years.

    I agree that the vent area calculations don’t seem to matter much. We work in many older houses with no soffit vents and small gable vents, or houses where the gable vents are partially blocked off, or where ridge vents were installed without cutting the plywood or the felt, so they don’t actually vent at all. All of these attics perform fine, in terms of moisture management. In this climate, a very small amount of ventilation seems to be perfectly adequate. In fact many houses around here have bath vents terminated inside the attics, and I have seen a grand total of zero problems from this detail out of hundreds of fans. I know this is not the case in colder climates, but it does seem to be the case here with 4-5000 HDD.

    Also, it is certainly true that effective ventilation does not solve some problems. For example, we remodeled a house once, and the insulation pieceworkers didn’t have enough vent channel with them, so they left it out of a cathedral ceiling area which consequently had little air movement. We had a cold winter a couple of years later, and the plywood started warping and expanding enough to see it through the shingles. We removed the plywood, found it soaking wet; we fixed the venting to ‘solve the problem’, and put it all back together. The next winter we found we had the similar problems (though much less severe). With the help of Alex Moore, one of the foremost weatherization pros in the US (rest in peace Alex), we found a number of bypasses and air leaks. In particular, a new humidifier we had installed on an old HVAC system with leaky ducts was pumping tons of humidity into the cavities. Effective air sealing on the ductwork and bypasses ultimately solved the problem, not correcting the ventilation. So I concur that no amount of ventilation will make up for major moisture sources, these should be prevented or fixed.

    There are three points where I would put a different slant on the information you provide.
    First, I know of several houses where roof longevity is very different on unvented roof areas versus vented roof areas, including the house I’m sitting in right now—my own—where the original house has gable vents but a garage addition has no ventilation at all. The shingles on the garage have failed (curled and lost their granules) while the ones on the house are fine. The house section actually gets more solar exposure and the other details of the roofs are similar (pitch, sheathing, etc.), so the ventilation is the only factor that would account for the difference. As I say, I have seen this on several other houses as well. I have no doubt that ventilation helps with shingle longevity, even if other factors also count.
    Second, in our hot summers, ventilation can make a big difference in comfort on the top floor. Many clients have told me they made a big change in their houses by adding power fans or effective (baffled) ridge vents. I agree that the fans probably use as much or more energy as they save, but energy savings is not what these clients wanted, they wanted comfort. We don’t recommend power vents because we’ve seen them depressurize the whole house and cause massive infiltration, moisture, and comfort problems in houses, and because proper air sealing and insulation are about as effective (not to mention correcting lousy ductwork). But there is no denying that well-vented attics are much cooler, and that can make a big comfort difference in some houses.
    Third, in my observation, all roofs leak. Period. Most types of roofing will leak a bit of water under some conditions (windy storms, ice or snow, etc.). Also, most people only repair or replace roofs after they see water inside. Therefore roof assemblies should be built with a high drying ability whenever possible, as they are much more likely to last without major problems. For us, even spray foamed roofs should be vented for this reason. We ask the insulators to put in vent channels, lapping them so any leaks run down toward the soffits. While this is not a perfect setup (as you mention, ideally there would be more space than vent channels create), in my opinion it is much less vulnerable to inevitable small or intermittent leaks, compared to an unvented assembly.

  19. GBA Editor
    Martin Holladay | | #19

    Response to Douglas Horgan
    All other factors being equal -- they rarely are -- a ventilated roof will be slightly cooler than an unventilated roof. Whether the difference in temperature will affect the longevity of the shingles is a matter of debate. (It is certain, however, that shingle color matters more than ventilation.) Some experts estimate that the shingles on a vented roof may last two years longer than the shingles on an unvented roof. This has not yet been proven, but if it matters to you, by all means vent your roof.

    In any case, most asphalt shingles are replaced before they wear out -- often because of unsightly algae growth.

    If your garage shingles are curling and have lost their granules, my first questions would be: (a) are you sure that the shingles are the same age as those on your house, and (b) are you sure that the shingles are the same brand?

    If clients are made uncomfortable because of high attic temperatures -- in other words, if their ceilings are getting hot during the summer -- I can only conclude that their ceilings have lousy insulation. If a ceiling is airtight and has code-minimum levels of insulation, properly installed, then the ceiling is never going to get hot enough during the summer to make the occupants uncomfortable. If homeowners told me that their ceilings were radiating so much heat that they were uncomfortable, my first remedy would be to improve the insulation above the ceiling.

    On the third point, I agree completely. Roofs leak, and it's always good to be able to inspect the underside of the roof sheathing. That's why the best arrangement is an unconditioned attic with insulation on the attic floor. Every roofer prefers to be able to climb up in the attic and inspect the roof sheathing. The best homes will have no cathedral ceilings.

  20. user-1140531 | | #20

    Cold-Side Ventilation
    With a cathedral ceiling, I feel that a bit of backup redundancy in removing any vapor that happens to get into the insulation cavity is prudent. It also seems prudent to remove any small amount of heat loss through the insulation that might develop ice dams. In the future, I will use metal roofing, so I am not worried about solar heat gain diminishing shingle life. With sufficient insulation, the problem of living space summertime heat gain could be overcome, but again, a bit of backup in the form of ventilation seems prudent.

    Since there is a limit to the practical effectiveness of air sealing in the walls, and because that limit can justify the backup measure of adding insulation outside of the sheathing, it seems equally prudent to vent cathedral ceilings because the same limit also applies to the effectiveness of making the ceiling airtight.

    I also disagree with the generalization that "The best homes will have no cathedral ceilings."

  21. GBA Editor
    Martin Holladay | | #21

    Response to Ron Keagle
    I agree with you that a ventilation channel provides more safety for cathedral ceilings. That's what I wrote in Comment #11: "When it comes to cathedral ceilings, a ventilation channel is much more important than it is for an attic. ... [For a cathedral ceiling,] I think that a ventilation gap is a good idea, because it can make the difference between a roof assembly that works and one that doesn't."

    It's OK if you disagree with me on the advisability of cathedral ceilings in general. There are plenty of people on your side of the fence (the pro-cathedral-ceiling side).

    In general, cathedral ceilings are risky, and they make inspection of the roof sheathing impossible. That said, it isn't always possible to put an attic under every roof. If there is a situation where a cathedral ceiling is unavoidable, there are ways to detail cathedral ceilings so they work. But in my opinion, an unconditioned attic is always preferable.

  22. doug_horgan | | #22

    We agree...mostly
    I think we're mostly on the same page, just making some limited, finer points about my climate.

    "If homeowners told me that their ceilings were radiating so much heat that they were uncomfortable, my first remedy would be to improve the insulation above the ceiling."

    This is absolutely what I recommend. It makes so much more sense than adding a fan.

    However, in my opinion it's better for professionals to be aware that those fans do help with summertime comfort in some situations. When it's time to have that conversation with the client, you don't want to be telling them "it doesn't help anything" when it often actually does, and the clients may even have direct personal knowledge that it does.
    Insulation, duct sealing, and attic air sealing help more, work all year round, and pay for themselves, which is why they make more sense. But, venting or fans do actually achieve some limited goals, at a high cost, so let's be aware of that when it's time to talk about it.

  23. GBA Editor
    Martin Holladay | | #23

    Response to Kevin Dickson
    My take on whole-house fans is a little different from yours. I think that whole-house fans can be an excellent way to cool a house in a dry climate -- especially if you live somewhere where temperatures usually cool off at night.

    The key is to buy a whole-house fan with motorized, insulated shutters. The best whole-house fans are manufactured by Tamarack.

    Whole-house fans have nothing to do with attic ventilation, as you accurately point out. Their purpose is to ventilate the house under the attic, not to ventilate the attic. I have written an article that tries to untangle the confusion some people have between whole-house fans and powered attic ventilators. Here is the link: Fans in the Attic: Do They Help or Do They Hurt?

  24. kevin_in_denver | | #24

    Whole House Fans Work great in some climates
    Whole house fans are usually installed in the ceiling plane and blow upward into a well-ventilated attic. Their main function is to pull cool summer night air into the house through open windows. For this reason they are often wrongly called attic fans. Installed this way, however, they sure do vent the attic!

    Since it's difficult to make them airtight, they have no business in a tight well-insulated home. They are also very difficult to insulate. But many thrifty folks in the Southwest love them if they don't have AC.

    Another good reason not to install them in a well-insulated home is because they won't cool during the day. Therefore, you probably want air conditioning. If you install, say, a 24 SEER minisplit, then you can't save much money by running a whole house fan at night. If you can't afford a minisplit, then a cheap swamp cooler won't cost much more than a whole house fan, and at least works during the day.

    But that's a whole nother topic.

  25. kevin_in_denver | | #25

    At least consider a swamp cooler
    A 1450 CFM Tamarack fan is $429, but a 2800 CFM evaporative cooler is only $286

  26. GBA Editor
    Martin Holladay | | #26

    Response to Kevin Dickson
    I agree that evaporative coolers often make sense, especially west of the Rocky Mountains.

    GBA has a section on evaporative coolers in our encyclopedia: Fans and Natural Cooling.

  27. SanDiegoSun | | #27

    Fire prone areas
    I'd like to see an article that addresses the issue of attic ventilation in California and other fire prone areas where soffit vents are prohibited in new construction due to the risk from wild fires.

  28. GBA Editor
    Martin Holladay | | #28

    Response to Jane Babin
    As far as I know, soffit vents are not prohibited in California, but the building code may require soffit vents to resist the intrusion of burning embers. If you live in an area with such restrictions, you can design an unvented attic, or can specify soffit vents that meet code requirements.

    Here is a link to a product that claims to meet code requirements for use in fire-prone areas:
    Vulcan Vent.

  29. regburns | | #29

    All About Attic Ventilation
    Fantastic article!! Very well written! Thank you Martin Holladay. I'll be passing this article around to all my production and sales staff, as you nailed a lot of the biggest misconceptions about ventilation.

    Reg Burns
    Cherry and Clark Roofing
    Mississauga, On

  30. Jay Walsh | | #30

    Gable vs Ridge Venting
    An observation I made a couple winters ago when we (here in Western Mass) were having heavy snowfall was that many homes with ridge vents, which were covered with 8-12 inches of snow for several weeks all has significant ice dams and ice-cycles. While many (most) those with Gable venting did not. While internally many other issues could be occurring, I concluded that with the airflow blocked at the ridge vent by snow accumulation, this stopped the attics from venting and the leakage of heat from the homes at the exterior wall plates (where insulation is least and many times airflow blocked) was causing snow melt resulting in the ice dams and ice cycles. Whereas the houses with the gable vents, which are always clear of snow, were functioning as designed and venting the accumulation of attic heat, from the airflow though the soffit vents and out the gables. Any thoughts on this? Keep an eye out for this and see what you conclude.

  31. user-1140531 | | #31

    Snow blocking ridge vents
    I have wondered why the issue of ridge vents being blocked by snow does not seem to concern many people. Maybe it is due to the possibility that many people believe that attic venting is only for cooling during summer. That would be the naturally intuitive conclusion. The reasons for wintertime venting are not so obvious. I have worked out a design for tall ridge vents to address the snow blocking issue.

  32. GBA Editor
    Martin Holladay | | #32

    Response to Jay Walsh and Ron Keagle
    Jay and Ron,
    There are three reasons that snow above ridge vents doesn't concern me:
    (a) Snow is not an air barrier;
    (b) Because of wind, thick snow doesn't stay on a ridge for long;
    (c) Attic ventilation isn't very important and is often counterproductive.

    Attics in adjacent houses are rarely identical.

    Some attics have ductwork; others don't.

    Some attics receive tremendous amounts of escaping heated air from the leaky ceiling below; others don't.

    Some attics have thick layers of insulation on the floor; others don't.

    If you are walking along a sidewalk looking at roofs in winter, you can sometimes deduce what is going on in the attic. But often, defects are hidden, and the causes of an ice dam are uncertain to the pedestrian.

    One thing is for sure: years of experience have taught home-performance contractors that you can't solve an ice damming problem by improving attic ventilation.

  33. Greg Labbe | | #33

    Amen & The resilience of wood planking!
    Mr. Holladay,

    Thanks for the thorough treatment this topic deserves! I appreciate your research and summary of this hot-potato topic. It confirms everything I've experienced in the hundreds of attics I've had the pleasure to be in and snoop around. I wrote a piece for Home Energy some time ago about roof topography, which looks at roof shape and acreage as a contributing factor too:

    One artifact/anecdote I've been ruminating on over the years is the benefit of solid planking for sheathing vs OSB/plywood. I've seen 150 year old wood plank look great - even when wet - seems to be more resilient than OSB/PW. I've seen OSB/PW replaced after only 3 years of service because ventilation was increased, but air leakage from living space ignored. The dimensional lumber planking with expansion gaps between planks seems more vapour open than the more modern alternatives.

    I just wanted to urge anyone with solid lumber sheathing to not be bullied into having to replace it as part of an ice dam remediation!

  34. user-1140531 | | #34

    Ventilation and Ice Dams

    Regarding your statement:

    “One thing is for sure: years of experience have taught home-performance contractors that you can't solve an ice damming problem by improving attic ventilation.”

    At first I was going to say that I disagreed with that, but when I parse the words, I am inclined to agree. The meaning pivots around the phrase, “an ice damming problem.” One might say that any ice dam is a problem; but probably the most common way of looking at it is that no ice dam is a problem until water starts coming through the ceiling and running down the walls. For an ice dam to get to that point of being a “problem,” there must be a major amount of heat and heated air escaping into the attic. I agree that an ice dam of that magnitude is not going to be easy to solve with ventilation, especially when the ridge vents are plugged with snow.

  35. ptombarrett | | #35

    Attic Venting
    Martin, great article. I did an attic venting research project here in California back in the 1990s and came to the same conclusions for the same reasons. Since then I've battled many "building experts" who claim attic venting is necessary to cool attic spaces but when I tell them about my study they can't get their heads around the facts. I monitored attics and interiors of homes with temperature probes and heat flow sensors and complied months of data with dataloggers and found no heat transfer with a well insulated attic deck even when attic temperatures hit 160 - 170 degrees.

    We use one home to experiment on various venting strategies typically found in California and found very little cooling or air movement effect from venting.

    One of the interesting things about heat buildup in attics and asphalt shingle warranties tied to the amount of attic venting needed to keep the attic cool, is that I recorded temperatures immediately under shingles between 180 - 190 degrees on south facing slopes and just above ambient temperature under wood shingles. The problem with heat is the shingles not the venting. White shingles are better, but the asphalt matrix gets hot and transfers that heat to the attic. It is a great deal for the shingle manufacturers when their shingles fail because of the heat they can attribute it to inadequate venting and void their warranty.

    Now can we get the building code changed to reflect reality?

  36. GBA Editor
    Martin Holladay | | #36

    Response to Thomas Barrett
    Anyone who looks closely at the way asphalt shingle manufacturers introduced attic venting requirements into their warranty rules reaches the same cynical conclusion about the manufacturers' motives.

    Suffice it to say that the manufacturers' attic venting requirements are not evidence-based.

  37. user-1140531 | | #37

    Venting Affect on Shingle Warranties
    From what I understand, it is generally believed that heat does shorten shingle life, but only by a couple years or so. Therefore this type of failure would be well beyond the warranty time frame. If failures occur within the warranty period, the manufacturer would have to show that the ventilation was insufficient to meet their warranty in order to deny a claim. In such a case, if the ventilation is in fact insufficient, why would one conclude that the cause of shingle failure was not due to insufficient ventilation?

    If shingles within the warranty period fail on a roof lacking adequate ventilation according to the manufactures requirements; and if the manufacturer denies the claim because they say the lack of ventilation caused excess shingle temperature; is the only alternative explanation that the shingles would not have been at a lower temperature even if ventilation was adequate according the manufacturers requirement?

  38. user-730247 | | #38

    About those shingle warranties
    Please raise your hand if you have ever even *heard* of a successful shingle warranty claim on a roof that did not have **serious** visible defects when the bundle was opened. Anybody? Anybody at all??
    I'm not saying it couldn't happen, but P=<0.05. That old argument for attic ventilation is right up there with "Eat your broccoli or you won't grow up nice and tall."

    This is an *excellent* article, and anybody who has anything to do with residential construction, renovation, or maintenance should be required to pass a test on its content. Thanks, Martin!

  39. user-1140531 | | #39

    Moisture & Ventilation in Photo Image 2 of 2
    In image 2 of 2, the caption says that this is a well-ventilated, well-constructed, relatively new building. The caption also implies that this is an example of a moisture problem in a ventilation zone, thus proving that ventilation is not capable of mitigating moisture, and therefore a reason why ventilation is unnecessary.

    But where is this moisture coming from?

    I would submit that this is not a well-constructed building, as the caption says, in that it has serious air leakage from the conditioned space into this attic space; and that the visible moisture is from the condensing of humid air leaking in from the heated space below.

    I would not expect ventilation to solve this problem. The article itself confirms that ventilation should not be expected to solve a moisture problem. Specifically, for example, Bill Rose says, “Don’t rely on ventilation alone to take care of moisture in the attic. The best protection against condensation and mildew in the attic is a dry basement or crawlspace. Also important is an airtight ceiling.”

    It is probably true that the darkened zone of wetting is the coldest due to the incoming ventilation outdoor air. But if that much water vapor is available through air leakage from the heated space, it will find a condensing surface in areas other than just that incoming air zone. I conclude that this mold problem would be much worse if no ventilation were provided.

    Overall, I would say that image 2 of 2 is an example of why cold-side ventilation is necessary as a backup feature to offset the difficulty in achieving 100% effective air sealing.

  40. GBA Editor
    Martin Holladay | | #40

    Response to Ron Keagle
    I wrote, “This photo shows sheathing mold near the ventilation baffles in the attic of a well-ventilated, well-constructed, relatively new building in the Pacific Northwest.”

    You wrote, “I would submit that this is not a well-constructed building, as the caption says, in that it has serious air leakage from the conditioned space into this attic space; and that the visible moisture is from the condensing of humid air leaking in from the heated space below.”

    It is possible that you are completely familiar with this building in British Columbia, and that you know more about the building than the researcher who took the photo. Possible, but not probable.

    I will quote from the paper where the photo was published (“Highly Insulated, Ventilated, Wood-Framed Attics in Cool Marine Climates,” by Patrick Roppel, Neil Norris, and Mark Lawton): “The attic venting area, including both intentional and unintentional openings, is presented in Table 2 for the measured pressure differential and is compared to the applicable required venting area required by code for the construction of the test attics (1:300 per insulated ceiling area). The measured venting areas are higher than the average areas reported in the BLP (1991) and Sheltair (1997) studies and exceed the building code requirement by as much as three times.

    “The measured air leakage area for the attic ceiling, calculated normalized leakage area (NLA), and observations of the smoke test are summarized in Table 3. These values were derived using equation 43, Chapter 16 Ventilation and Infiltration, of the ASHRAE Handbook—Fundamentals to convert to a 10 Pa pressure differential basis. Comparison of the NLA values in Table 3 to the reported values in past Canadian studies (BLP 1991; Sheltair 1997; NRCan 1997) suggests that the ceiling airtightness of the units in this current study can be considered to have at least average airtightness levels, by Canadian standards, and a convincing argument can be made to classify the ceiling to attic interface as airtight.”

    You asked, “Where is this moisture coming from?” The researchers concluded that this is exterior moisture. The researchers observed and photographed similar mold on the underside of roof sheathing on roofs over exterior mailboxes (like the roofs you see over a picnic table at a state park).

  41. user-1140531 | | #41

    Response to Martin Holladay


    Thanks for your explanation. You are correct that I have never seen the building in the photo. I concluded that the moisture was coming through air leaks from the conditioned interior. I had not considered that the moisture could be coming from the exterior, as you say was concluded by the researchers.

    I am curious about the mechanism for delivering this exterior moisture into the attic space and depositing it on the underside of the roof sheathing. Is this from the intrusion of rain or snow being drawn into the eave vent openings? Or is it from water vapor being drawn into the eave vent openings and then condensing on the underside of the sheathing?

    If the latter, I don’t see how the underside of the roof sheathing would be a condensing surface for exterior vapor in either winter or summer-- because in either season, the underside of the roof deck would not be colder than the outside temperature from where the vapor originated. So why would any feed of outside vapor condense in the attic?

    Did the researchers explain why wetting occurred on the underside of open roofs over mailboxes and picnic tables? I suppose there is a possibility that roofs could lose heat by radiation to deep space on cloudless nights, and thus become colder than the ambient air temperature. If that were the case, it might also explain outside vapor being drawn into eave vents and condensing on the underside of the roof deck.

    What did the researchers recommend as the remedy for preventing the wetting effect on the underside of the roof deck?

  42. GBA Editor
    Martin Holladay | | #42

    Response to Ron Keagle
    The outdoor air in the Pacific Northwest is very humid. When the air is humid for much of the year, exterior wood work can get moldy (unless it is exposed to regular sunlight).

    The researchers concluded that, in the climate of the Pacific Northwest, moldy roof sheathing in vented attics may be unavoidable, and we may need to educate real estate agents (telling them that mold on attic sheathing is not a health risk).

    There are several possible reasons that the mold shows up near the soffit vents. It's possible, for instance, that the upper parts of the attic stay dryer because those areas are warmer. It's also possible that the frequent streams of outdoor air in this part of the attic keep the sheathing damp in that area.

    1. ArchitectJudge | | #85

      Thanks for addressing this photo and the following questions Martin. I am in the PNW and am curious what one should do to avoid the moisture condensation on the roof sheathing. I am going to spec CDX plywood instead of OSB, but besides this what could be done?

      Should I expect to have to replace the wood whenever I re-roof the shingles (20+ years)? I supposed that is not that big of an issue in the grand scheme of things, but curious what the best practices are. I will be doing 1/150 calc for roof ventilation, and 60% of venting at eves and 40% at ridge

      1. Expert Member
        MALCOLM TAYLOR | | #86


        Since Martin wrote this in 2013 there has been research by RDH which attributes the moisture problems in PNW roof sheathing mainly to Nigh Sky Radiant Cooling. He wrote an article in 2017 exploring the subject, which I think you will find interesting.

        My own experience with roofs here on Vancouver Island is that while the sheathing on a well air-sealed and properly vented roof may exhibit a bit of mold or discoloration, in the absence of other problems it does fine, and should not need replacing when the shingles do. I've worked on a lot of roofs constructed as you are proposing that are 40 to 50 years old where the plywood sheathing looked much as it did when it was first installed.

        1. ArchitectJudge | | #87

          Awesome, thanks for the updated information and for setting me at ease. Much appreciated!

  43. user-2589775 | | #43

    additional bracing for walls
    Years ago, when my father was custom building houses and foam sheeting was new to the market . we added metal strapping(Banding as is used on loads of lumber) on the diagonal to all exterior walls both interior and exterior. The banding was self tightening in that driving an 8d nail into each stud tightened and tensioned the strapping. This was done without an engineers advice but could easily be used underneath drywall without causing lumps and bumps and was very inexpensive. It seemed to add a great deal of rigidity to the assemblies. The cost was negligible when purchased on the bulk spool. Any thoughts?

  44. GBA Editor
    Martin Holladay | | #44

    Response to Thomas Baldes
    I'm not sure what your question has to do with attic venting -- perhaps you accidentally posted the question on the wrong page -- but I'll be glad to respond.

    Many builders use diagonal steel strapping to brace walls. Rather than using a banding machine, however, there is an engineered solution: you can use Simpson Strong-Tie’s L-profile strapping (the product designation is Simpson RCWB). For more information, see Four Options for Shear Bracing Foam-Sheathed Walls.

  45. Logger6644 | | #45

    All About Attic Venting
    I've been building my own log home with a cathedral ceiling and so have been following this and related blogs for some time. My ceiling is currently a cold metal roof with a 3" vented air space above 1/2" plywood covered with 30# felt. The 2x12 rafters are filled with r-38 fiberglass. I was planning on covering that with t&g boards until I reviewed Martin Holladay's July 2012 article in Fine Homebuilding. I now plan on the vented roof design with the rafters being capped with the rigid foil faced foam with taped joints and the t&g fastened to strapping installed on top of the foam. The only material difference is the venting is on top of the roof sheathing instead of below. Is this a viable design or do I need to change something to improve it?

  46. GBA Editor
    Martin Holladay | | #46

    Response to Michael Koelsch
    It's perfectly acceptable to put a vent channel above the roof sheathing.

    It's important to pay careful attention to airtightness with this type of roof assembly, however. If you are depending on a layer of interior rigid foam to be your air barrier -- and it sounds like you are -- you need to do a meticulous job of sealing the perimeter of each sheet of foam. It's also important to seal all penetrations.

    For more information, see How to Build an Insulated Cathedral Ceiling.

  47. Logger6644 | | #47

    Cathedral Ceiling Insulation
    Thanks Martin, as I've undertaken building this house as a ten year and counting retirement project with my wife and I doing everything from digging the foundation to building cabinets from lumber on the property, I've relied heavily on experts like yourself and others to avoid catastrophic errors.

    As you suggest, I will do my best at achieving a six sided air tight seal in every rafter space. I'm glad I did not rush the finishing of the ceiling until I watched it for a couple of years and did notice it lack of performance for the reasons you outline. Thanks again for the service you provide and it's obvious benefits for us and the planet.

  48. Kamboji | | #48

    Unvented Cathedral Ceiling?
    Thank you for your very informative posts. We live in an older home (built 1915) near Bangor, ME. The attic is very large and we are planning to fix it up. Currently we have very old fiberglass (R-11) and rockwool on the high cathedral ceiling. We previously blew insulation into the floor. The plan is to rip out all the old ceiling insulation, reinsulate, and then add sheetrock, new flooring, and a little lavatory. I believe that there is no ridge vent and that the soffits are unvented. I'm not entirely sure about this, but I reached into the soffit beyond the floor, and I could not feel any outlet to the outside, and the space is filled with the blown insulation we added to our attic floor. The contractor suggested leaving a gap, then adding R-23 Roxul batts (our rafter beams are only 2x6), and then in the summer getting someone to drill holes from the outside into the soffit (this seems a little strange to me). If it turns out that there's no ridge vent as suspected, they propose adding gable vents across the top. After reading your posts, I am now leaning towards an unvented ceiling, something like this (in this order):

    -Add foil-backed polyiso (3-4 inches) right against the sheathing, foil side down.
    -Caulk the polyiso well along the sides
    -Add batts of Roxul up to the edge of the rafters.
    -add vapor barrier
    -Attach foam board under that to underside of rafters
    -Add sheetrock etc.

    I don't think this will get us all the way to the recommended R-49 for zone 6 but we don't plan on using the space a lot (only heating it when in use), and the floor is pretty well insulated.

    Does that sound reasonable? One more thing- we really do not want to use spray foam for health reasons. Thanks again for this valuable information!

  49. GBA Editor
    Martin Holladay | | #49

    Response to Hari Kamboji
    The insulation method your propose is called cut-and-cobble. For more information, see Cut-and-Cobble Insulation.

    I don't recommend the cut-and-cobble method for sloped roof assemblies or cathedral ceilings; read the article to learn about some recent failures using this method.

    If you want to insulate a sloped roof assembly, I recommend that you install at least enough R-value to meet minimum code requirements. You really do want to install R-49 if possible. If you are worried about head room, you can install some of your insulation in the form of rigid foam above your roof sheathing. Of course, that means that you will need to install new roofing.

    For more information on your options, see How to Build an Insulated Cathedral Ceiling.

  50. Kamboji | | #50

    Re: Response to Hari Kamboji
    Dear Martin,
    I had not seen your post on "cut and cobble"- very helpful. Thanks a lot. I got this idea after reading your post on cathedral ceilings and then reading the 2009 IRC code (R806.5.3 of the 2012 code still seems to suggest it). I wondered why the code seemed to invite cut-and-cobble for conditioned attics, while this was not mentioned in the conclusion of your cathedral ceiling article. Now it is clarified (there remains risk of vapor penetration).

    The idea of rigid foam above the sheathing is appealing. In that case, the Roxul can go right up against the underside of the sheathing, right? Our roofing was installed in 1997, so maybe it's getting due for replacement anyhow. But it actually still looks pretty good to my untrained eye.

    Head room is not a huge concern. I've seen mention of various techniques to extend the insulation space, but I haven't seen a dedicated explanation (e.g. what's "furring"?). Something like that, in combination with foam board atop sheathing could get us to R-49 or beyond.

    Thanks for your reply.

    [Editor's note: To read the response to this comment, and other comments that follow, go to page 2 by clicking the number 2 below.]

  51. GBA Editor
    Martin Holladay | | #51

    Response to Hari Kamboji
    "Furring" is a synonym for "strapping." It refers to pieces of lumber -- usually 1x3s, 1x4s, or 2x4s -- that are attached perpendicular to the framing (studs, joists, or rafters) or on one side of rigid foam (in which case the furring is usually parallel to, and directly above, the framing on the other side of the rigid foam).

  52. Kamboji | | #52

    vapor barriers inside and out?
    Because Roxul insulation is unfaced, I read that one needs to install a vapor barrier (or is it air barrier?) on top of it (facing the living space). On the other hand, if one applies foam board on top of the roof sheathing, that also acts as a vapor barrier. Is it bad practice to have vapor barriers on both sides of the roof? One website I found recommended using a permeable insulation on top of the sheathing (like heavy duty mineral wool batts) if a vapor barrier has been installed inside (I realize such an assembly would violate existing code).

    A somewhat unrelated question: It may be a year or two before I can redo the roof to add the foam board on top. Rather than cut-n-cobble, I'm planning to go ahead now and install the air-permeable Roxul insulation between the rafters up against the underside of the sheathing (an unvented setup), then vapor membrane, then polyiso boards across underside of rafters, then sheet rock. To mitigate condensation, until I redo the roof, I will just keep the attic unheated (there's insulation in the floor). Our house uses forced hot air heating, so the interior humidity is very low in winter. Comments are welcome if anyone thinks this is a terrible plan. The house has already had fiberglass and rock wool between the rafters for 50+ years without any moisture problems arising, so this would seem a safe thing to do in the short term. Thanks again!

  53. GBA Editor
    Martin Holladay | | #53

    Response to Hari Kamboji
    You are correct that you don't want to enclose your Roxul with a vapor barrier on both sides. If the Roxul is part of a vented roof assembly, you usually want to install a vapor-permeable air barrier above the Roxul; this is your ventilation baffle. Thin plywood or stiff cardboard will work; so will thin EPS, which is fairly permeable to water vapor.

    Underneath the Roxul, facing the living space, you also need an air barrier. In most cases, the ceiling drywall is an air barrier.

    I don't recommend installing Roxul in an unvented roof assembly unless you are also installing rigid foam above your roof sheathing. If you proceed with your plan, your roof sheathing could accumulate moisture and develop rot or mold.

  54. Kamboji | | #54

    Many thanks for the quick
    Many thanks for the quick reply. Just to clarify, in a vented assembly, a vapor barrier should be applied on the underside (living space side) of Roxul, but in an unvented assembly which includes the rigid foam on top of the roof sheathing, one should not apply a vapor barrier anywhere on the interior side?

  55. GBA Editor
    Martin Holladay | | #55

    Response to Hari Kamboji
    You don't want an interior vapor barrier. In some climate zones, building codes require a vapor retarder -- which is a less stringent layer than a vapor barrier.

    In general, there aren't usually any reasons to worry about vapor retarders in ceilings or walls (with a few exceptions). In most cases, the only reason to install a vapor retarder is to satisfy your local building inspector. If the inspector wants one, I suggest that you use vapor-retarder paint or MemBrain (a smart vapor retarder).

    While vapor retarders are usually unnecessary, and vapor barriers often cause problems, an interior air barrier is always a good idea.

    For more information, see:

    Vapor Retarders and Vapor Barriers

    Forget Vapor Diffusion — Stop the Air Leaks!

    Do I Need a Vapor Retarder?

  56. Kamboji | | #56

    polyiso under rafters as vapor barrier
    Thanks- as always these links are extremely informative, and I'm sorry for not stumbling upon them before asking. In the comments of the first link above you mention that foil-backed polyiso acts as a vapor barrier. Does this mean I should not be attaching it below the rafters for extra R-value?

  57. GBA Editor
    Martin Holladay | | #57

    Response to Hari Kamboji
    Interior polyethylene can sometimes cause problems in the summertime, because it reduces inward drying and (if the house is air conditioned) can get cold enough to be a condensing surface for any moisture in the framing bays.

    Polyiso won't ever get cold enough to be a condensing surface, even if the house is air conditioned, because it has R-value. It still reduces inward drying, however, so walls or ceilings with interior polyiso should (ideally) be able to dry to the exterior.

  58. energybloke1 | | #58

    attic ventilation
    As an energy consultant working in a cold climate (7500 HDD), I have had the opportunity to investigate many home performance problems to include ice dams and condensation issues inside attics and have quite often found and photographed soffit vent screens that were plugged or at least partially plugged with airborne "debris" that rendered the vents to be less than effective, or worse. As an instructor of building science basics, I've asked hundreds of participants (mostly remodelers) how often they make the effort to clean soffit vents. Rarely does anyone even acknowledge the need to do so. In fact, when I mention the notion of doing so, they tend to look at each other as if to convey an attitude (collectively) that I must be from another world. In addition, I've seen and have photographed roof vents semi-plugged with insect nests. I feel confident in assuming the same reduction of airflow happens with ridge vents over time. My point is passive ventilation can deteriorate over time and cannot be depended upon to a permanent working system. Even passive ventilation systems need maintenance. Perhaps a better solution is designing and constructing unvented roofing assemblies.

  59. GBA Editor
    Martin Holladay | | #59

    Response to Kevin Hogan
    You raise an interesting point. You're right, of course, that soffit and ridge vents (just like the air intake vents for HRVs) can become clogged by spider webs, dried leaves, hay, and insect nests. If anyone is depending on the free flow of air through their soffit vents to keep their attic functioning well, I guess they need to get a vacuum cleaner that they can use while standing on a ladder.

    Here's the main point of my article, however: if you are depending on attic ventilation through soffit vents to keep your attic dry, you've done a bad job of air-sealing your ceiling. If you build your house correctly, and pay attention to airtightness, you don't really need to depend on attic ventilation to keep your attic dry.

    When it comes to preventing ice dams, the two most important factors are (a) an airtight ceiling, and (b) a thick layer of insulation. Venting matters much less than these two factors. For more information on ice dams, see Prevent Ice Dams With Air Sealing and Insulation.

  60. rotary82 | | #60

    not enough venting
    I have an attic converted to living space where the roof is also the cathedral ceiling. Most bays go from soffit to ridge and are insulated with fiberglass and a foil barrier with a Duravent installed so that there is always an airspace from soffit to ridge. Those bays are mostly fine. The areas I'm having a ton of ice and moisture on the sheathing is where the bays are interrupted and there is only a ridge vent. Should I fill the bays solid with spray foam? Or should I at least foam up the ridge vent from the inside on those bays? If I do a perfect job with the foil barrier will that stop the moisture from getting into the bays? I have a damp basement and as much as I want to seal it up it is no way in my price range.

  61. GBA Editor
    Martin Holladay | | #61

    Response to Mark D'Amico
    It sounds like you have mold on the underside of your roof sheathing. The cause of this problem is almost always a leaky ceiling. These air leaks are allowing warm, humid interior air to enter your attic, and the moisture in the air is condensing on the cold roof sheathing.

    If you fix this problem by sealing the air leaks, there will be two benefits: (1) Your energy bills will be reduced, and (2) The mold will be eliminated.

    Here is a link to an article which explains what you need to do: Air Sealing an Attic.

    If you won't want to do this work, you can follow the route you suggested: "using a closed-cell spray foam directly on the underside of the roof." Done properly, that approach will also reduce your home's air leakage problem and protect your roof sheathing. However, that approach is significantly more expensive than sealing the air leaks at your ceiling plane. For more information on the spray foam option, see Creating a Conditioned Attic.

  62. markmdz | | #62

    typical 50's ranch - awful venting, moisture, poor insulation
    First, thank you for the informative article. Second, sorry to comment on such an old publication, but I feel compelled to do so and ask a few questions.

    The majority (read: all but the last statement) discuss the disadvantages, or perhaps the lack of advantages, to a vented attic at length. Despite that, the last statement is recommending a vented attic done properly. I now have a better understanding of the attic venting and so many of the misconceptions or inaccurate claims associated.

    So why not just seal it right up start to finish?

    I have a small ranch in New England which was built in the 50's and has, as anyone would expect of a house built in this era sub-par insulation. When it was built, the house had two small gable vents, which it still has. About 20 years ago, the roof was replaced and a ridge vent was added at that time. There are currently no soffit vents.

    The attic floor was given an R-19 treatment between the ceiling joists. This was a half-hearted attempt by a previous owner, as it's incomplete. Perhaps it's still better than nothing.

    I had the roof replaced a few months back and again the roofer recommended a ridge vent "because it was there with the old roof, and any ventilation is better than none".

    So, this brings me to the question again of what are the benefit and disadvantages of using a closed-cell spray foam directly on the underside of the roof? [Of course the first step to this would be to close or remove the gable vents and seal off the ridge vent.] The way I see it, it'll increase the conditioned space by including the attic cubic footage, but it's a much better insulation than the R19, which once the air escapes that, it's gone. So perhaps the added R value and better sealed 'envelope' created by the spray foam, despite the added volume of 'conditioned space', nets out to a benefit over the R19 on the attic floor? I don't know.

    Then of course that brings to light the argument that the reason the underside of roofs are not painted, nor should they be insulated, is to allow the wood to 'breath' and not trap moisture in the wood. Is this as much of a concern as some claim? Is it flawed thinking that whatever moisture is in the wood is really negligible and not a concern of allowing it to 'breath'? To further that, my thinking (and very likely flawed) is that the spray foam benefits (aside from the insulation factor) would be: Once applied, whatever excess moisture is in the wood will be extracted by the foam during the curing process (as the foam needs moisture/humidity to cure) and it'll seal the wood surfaces from direct exposure to moisture and significantly reduce the amount of exposed organic surface material (i.e. wood). Is this all bunk, or is there some validity to this argument?

    The last byproduct of an insulated roofline rather than heavily batted attic floor is the new found availability of semi-conditioned storage space.

    So after all that, given there's already R19 in there, do I just go buy a bunch more of the r19 and lay it over the existing stuff and call it a day?

    Please help shed some educated light on this mind-bending question of what's best.

  63. Dana1 | | #63

    "...significantly more expensive..." indeed!
    A square foot of R20 closed cell foam costs about $3-3.50, and even in zone 5 (southern New England) you need at least R20 for dew point control with an R49 code-min stackup.

    A square of R20 cellulose costs about $0.60, and the attic floor has fewer square feet than the roof deck. To hit code min R49 you'd have to add ~R35 ( at about a buck-a-foot) over the low density R19s, which will get compressed a bit by a 10" overburden of cellulose, Target total depth is 15-16" to allow for some amount of settling over time down to 14".

    If you have enough space to increase the total insulation depth to 15", after air sealing at the ceiling plane blow cellulose over the existing R19s and you'd hit current code-min. But even if you can only bring it up to 10" that's still a huge improvement. Make a bunch of paper/cardboard depth gauges to keep track of it, making it smooth & even depth everywhere. You'll still have to keep the insulation an inch away from the roof deck at the eaves- use the purpose-mad de asphalted cardboard chutes for that, not the cheap thin foamies. Open blown cellulose is a fairly easy DIY with a box-store rental blower (usually free rental with a minimum purchase of cellulose.)

    An overlayer of R19s is a much worse performing solution, due to the very high air permeance of the low density batts. R19s are more of an air-filter than an air-retarder, and take a serious performance hit unless there are air barriers on BOTH sides. Cellulose is about an order of magnitude more air retardent than low density fiberglass, and performs well even without a topside air barrier. As little as 3" of cellulose overtopping will "restore" the performance to the R19s, but it's worth installing more than that in any New England location.

  64. MrPeteH | | #64

    Venting challenges in Colorado
    Hi! This has been an extremelhy helpful article. I have never had trouble with attic venting until we purchased a brand new home at 7000' in Colorado, with a vented attic. Many may not be aware that even "in town" here, a storm with 40-50mph wind and 60-70+mph gusts is not that unusual. Gusts in the 100+ and even 200mph range have been measured. Builders must be careful to finish sheathing homes quickly after they begin! :)

    After a Nov '15 blizzard, we experienced several of the issues you describe (including snow/ice injected directly into the middle of the insulation due to improper perimeter barriers.) I've retrofitted with proper barriers/baffles (Accuvent worked nicely).
    The good news: the new barriers work. The insulation behind the barriers stays in place and is dry.

    The bad news: in last week's blizzard, while no snow/ice was injected into the middle of the insulation, the nice big ventilation channels allowed significant amounts of snow/ice to still enter the attic at the leading edge of the storm, and create snow piles on top of the insulation... enough that I couldn't remove it in time, we had ceiling leaks in our new home, and a full 4x8 sheet of ceiling is in precarious condition after getting wet.

    I have some ideas and questions, and am wondering what you think.

    a) Even though the storm came from the west and north, we ONLY had snow enter in the NW corner. b) This corner sticks out about 8' from the home next door; we're on the side but at the downwind end of a culdesac, with no homes (just a gentle hill) downwind of us for several hundred feet.

    Idea: I'm thinking that
    a) the street and especially home only 10' away essentially "focused" the wind.
    b) The air had to go somewhere, and due to Bernoulli effect, it likely hits the exposed wall of our home at higher than normal pressure (Florida Building Dept work on Wind-Driven Rain research estimates pressures easily reaching 2x or 3x normal
    c) Thus, that corner of the home may be exposed to far higher wind speed than anticipated, causing snow-laden wind to easily enter the attic.

    Question: Does this make sense or am I just blowing smoke?

    My thought: now that I've properly blocked the edges, use waterproof spray-in insulation foam to seal the baffles in that corner, so that any snow or rain trying to enter will melt and fall back out into the soffit area.

    Question: does that make sense?

    (I do not have the equipment to reach the soffit in that corner externally. It is 15'+ above a steep riprap slope.) My home IS still under construction warranty... so far the builder has mostly said "hey, it passed code inspection so what do you want?" ... although they HAVE agreed to blow in replacement insulation after I'm done properly repairing the blocks/baffles. TBH, I don't trust the contractors to fix it right after I've seen other work they've done... I am MUCH more aware now than a few months ago, thanks to you!

    Final question: do you know of any research or rules of thumb for preventing wind-driven snow/rain issues in a properly blocked/baffled attic?

    We are quite concerned. So far, we have been here when the blizzards have arrived. What happens if we are traveling?!

    Thanks MUCH. I am happy to invest more of my time helping all of us learn more. I have photos illustrating my own situation... even some accidentally taken during a framing walk-through. If I had known then, I would have pointed out the "blocks" made of loosely placed insulation batts. Completely inadequate of course.

    The interesting thing: in the November blizzard, EVERY home builder here got slammed. They are modifying their home designs and practices to help... yet still all have vented attics with baffles providing air flow... so I am not sure the problem described in this message is really being addressed.

  65. GBA Editor
    Martin Holladay | | #65

    Response to Pete Holzmann
    As my article explains, the fetish for venting attics is mostly a historical accident, rather than a necessity supported by building science. Builders and homeowners need to use common sense. If your vent openings are allowing snow to enter your attic, the obvious solution is that the problematic vent openings need to be sealed.

    If you ceiling is relatively airtight -- and I hope that it is -- no don't need to ventilate your attics. The only person who cares, really, is your local building inspector. If the inspector was satisfied, you can do whatever you want now that it's your home. (When it comes time to sell your home, however, you may need to discuss the status of your soffit vents and ridge vents again... but I suggest that you cross that bridge when you come to it.)

    In a very windy area with lots of cold, fine, powdery snow, roof vents can allow snow to enter the attic. I've seen it in Vermont many times -- but usually the amount of snow is smaller than the amount you describe.

    Do you know if your ceiling is airtight? Did your home ever have a blower door test?

    Assuming that your ceiling is relatively airtight, I suggest that you seal all of the problematic vents. Good luck in your negotiations with your builder.

  66. cabinflyer | | #66

    Just got some continuous
    Just got some continuous soffit venting from the lumber yard. Any reason the vent is only 94 1/2" long?

  67. GBA Editor
    Martin Holladay | | #67

    Reply to Steve Lenertz
    It's probably the same reason that a 1-pound can of coffee now contains only 13.5 ounces.

  68. SteveG107 | | #68

    Attic and roof used interchangeably
    Hey Martin,
    The terms attic and roof are sometimes used interchangeably.
    Can you clarify In this video what Matt Risinger and Joe mean by unvented roofs are the way to go?

    In all of the videos , articles and books I have read it is my understanding that in hot-humid climates you want the attic to be sealed (unvented) but the roof is to be vented.

    I am curious if building a home with a sealed attic and metal roof in a hot-humid climate needs the roof to be vented.

    I am seeing in many places the metal roof is installed on diagonal laths which essentially is venting the roof.


  69. GBA Editor
    Martin Holladay | | #69

    Resonse to User-7119046
    It would be easier to have this conversation if you told us your name. (I'm Martin.)

    Q. "Can you clarify in this video what Matt Risinger and Joe mean by unvented roofs are the way to go?"

    A. I have no idea what Matt and Joe were thinking, needless to say, and I don't want to be put in the position of speaking for either of them. In the video, Lstiburek said, “Vented roofs leak, vented roofs blow away, and vented roofs burn. Unvented roofs do none of those things. Duh.”

    From context, it seems that he could be referring to either (a) the ventilation channel in a cathedral ceiling between the top of the insulation and the roof sheathing, or (b) the ventilation channel under metal roofing between the underlayment and the metal roofing.

    Both types of ventilation channel can be omitted if you don't want the downsides listed by Lstiburek. Of course, if you are building an unvented insulated roof assembly, you have to follow the usual rules.

    I think that an air space between the roofing underlayment and the metal roofing is useful, because it catches condensation drips and allows these drips to evaporate. That said, such an air space is not required. I have no idea what Lstiburek's opinion on this issue is, but it sounds like he is recommending that this type of air space be eliminated.

    The purlins (1x4s or 2x4s) under metal roofing don't have to be diagonal -- in fact, if they are diagonal, roofers will curse, because it's tricky to know where to place the fasteners. These purlins can be parallel to the ridge -- the evaporation of the condensation drips will still occur that way.

  70. severaltypesofnerd | | #71

    What venting advice would you give specifically applicable to coastal California climates? Moderate everything: moderate humidity much of the year, moderate rainy season, 9+ months dry weather.

    I get why permeable tar paper under a metal roof is OK. But why is an impermeable "ice and water" type barrier OK right at the metal? Won't that keep any condensation moist for a long time?

    1. GBA Editor
      Martin Holladay | | #72

      The vapor permeance of your roofing underlayment is irrelevant, since (in most cases) there is no outward drying through metal roofing.

      In your climate zone, advice on whether to choose a vented or unvented approach is that same as in other climate zones. Either approach can work, as long as the details follow the usual rules.

  71. severaltypesofnerd | | #73

    @Martin the existing ceiling it not yet air sealed. The attic will remain unconditioned.
    The attic is unvented asbestos shingle. There are no eave vents at all.

    The metal roofers want to add a ridge vent. The question becomes how to make the whole system work best. Right now it probably works because the leaky ceiling allows lots of air flow, drying to the conditioned space. In the mild climate, no freezing, no humid days, probably works out OK.

  72. Harr1z | | #74

    Martin, Rehabbing a 1950's Cape Cod. Insulating the knee wall areas for storage and just for the additional insulation factor. Before reading this article, I was planning on filling the 2x6 rafters with: 1in air vent, 1 inch ridged polyiso board, 3.5 in unfaced fiberglass, then put 1in polyiso board over the rafters to seal it all in, to include the small slanted ceiling area which would then get drywalled. So about R25 not counting the R13-15 in the Knee wall framing. and second level floor joist. Now, Reading through the comments and your replies, I am curious if putting the fiberglass between polyiso boards is a bad idea (cheaper than all polyiso board). And, since I believe I can seal both ceiling planes pretty well, I dont need the venting here in SE Virginia, and would go 2inch polyiso under the sheathing. Good idea? The house still has the old 1x6 sheathing. Mostly in great shape. Thanks!

  73. GBA Editor
    Martin Holladay | | #75

    You can use 1-inch-thick polyiso as your ventilation baffle, and also include a continuous layer of polyiso under the rafters if you want; no need to worry about the two layers of polyiso causing any problems. That approach is described in the following article: "How to Build an Insulated Cathedral Ceiling."

    You're mistaken if you think that you don't need venting in Virginia. Unless you are either using closed-cell spray foam insulation under the sheathing, or an adequately thick layer of continuous rigid foam on the exterior side of the sheathing, a ventilation channel under the roof sheathing is required.

  74. george_master | | #76

    You mentioned that it is a “terrible practice” to locate HVAC equipment and ductwork in a vented attic. Does the same hold true for an ERV or HRV system with insulated duct work? I’m looking to install a balanced ERV system in my 1980s home in suburban Philadelphia (warm & cold climate) with a vented attic and about 12” of fiber glass insulation over the ceiling. The contractor said that this location was the most cost effective way to go with vents going into the ceiling and walls on the ground floor.

    1. GBA Editor
      Martin Holladay | | #77

      It's always best to install an HRV or an ERV inside a home's conditioned envelope. Running ductwork through an unconditioned vented attic can lead to several problems, including higher than usual energy bills, possible condensation in the ducts, and a cold HRV core or ERV core, reducing equipment efficiency.

      Remember, too, that an HRV or ERV requires regular maintenance, including frequent filter changes or filter cleaning. Easy access is essential.

      Some HRV or ERV manufacturers may not permit installing their equipment outside of the home's thermal envelope, so be sure to check with the equipment manufacturer before proceeding with your plan.

      For more information, see this blog by Allison Bailes: "Getting Mechanical Ventilation Inside Conditioned Space."

  75. d_w | | #78

    Hi Martin. Disclaimer: I am a homeowner with no building science training or background. I have been combing through various GBA articles, and they have helped me do what other websites have not: get a clear idea of what I don't know, in order to ask for help. Unfortunately, I am in a location with very few resources when it comes to green building or renovating.
    I hope you can offer advice about my unconditioned attic. Basics: 80-year-old house in Kansas, and I am nearly done removing the original rock wool insulation and fiberboard covering in order to air seal. The attic houses all the central air ducting for the second story of the house; I wish it were different, but it's not an option to move it or switch to ductless minisplits due to cost. The ducting is wrapped with fiberglass insulation. The current ventilation is achieved by gable vents and a central ridge square opening, no soffits nor continuous ridge vent. The roof was replaced 2 years ago. I plan to blow in cellulose after air sealing and an energy audit. Here are my questions:
    1. My impression from various articles on GBA is that attic ventilation is not usually a big concern, as long as there is no evidence of any problem. Does ventilation become a higher priority when ducting is inside an unconditioned attic? In my case, should I prioritize attic ventilation in order to reduce the attic's ambient summer temperature? Should I plan to blow in insulation to cover the ducts? Both? Neither?
    2. There is an unused brick chimney in the attic as well, with an unsealed chase going down to the basement. I read in the air sealing article from 2013 that removing the chimney is the best option. That is not part of my plan right now. Since the chimney does not communicate with a fireplace, but was rather for the since-removed boiler, it will not be used. Am I still restricted, by code or common sense, from the cellulose insulation directly contacting the chimney? If so, do you have a resource that will give me instructions on how to insulate around it? Or is the effort essentially a waste, and I should just pay to have the chimney removed below the attic floor?

    Thanks for the time and consideration.


    1. hockipuck | | #80

      Daniel, one thing to keep in mind with ducts in the attic is condensation on ductwork. While it may seem intuitive to pile cellulose or other insulation on top of ductwork in an attic to insulate it better, moisture still moves through insulation and can condense on the now colder ductwork. Two options I can think of, and that depend on the complexity of your duct runs: 1) use foamboard and either tape or acoustical sealant to create sealed chases around your ductwork. Seal this to the ceiling and then blow insulation over top. 2) Bow to the ease of spray foam in this instance and have someone spray 1.5-2" of CCF to encapsulate the ducts and then blow insulation around them.

      Either way, take the opportunity to make sure your ducts are sealed and insulated well. I would love for the experts to chime in with better options!

  76. GBA Editor
    Martin Holladay | | #79

    Q. "My impression from various articles on GBA is that attic ventilation is not usually a big concern, as long as there is no evidence of any problem. Does ventilation become a higher priority when ducting is inside an unconditioned attic?"

    A. No, although potential moisture problems in imperfectly vented attics can be disguised by the presence of leaky attic ducts. (Leaky attic ducts can help dry out attic air by heating the attic in winter and dehumidifying the attic in summer, assuming that the house has central air conditioning. Needless to say, this method of drying out attic air wastes a lot of energy.)

    My advice remains unchanged: If your attic has attic ducts, you should try to convert your vented attic into an unvented conditioned attic. For more information, see "A New Look at Conditioned Attics."

    Q. "Am I still restricted, by code or common sense, from the cellulose insulation directly contacting the chimney?"

    A. As far as I know, code requirements concerning insulation contact with chimneys still apply to unused chimneys (because the next owner of your home might decide to use the chimney.)

    Q. "If so, do you have a resource that will give me instructions on how to insulate around it?"

    A. Yes. See this article: "Air Sealing an Attic."

    Q. "Should I just pay to have the chimney removed below the attic floor?"

    A. If you can afford the cost of the work, that would certainly be the best approach.

  77. davehica | | #81

    Hi Martin - Love the article and exchange of ideas on GBA. Hoping you can help me before I spend a significant amount of money doing something incorrectly....

    I live in northern california in a 1917 craftsman bungalow. I implemented the corrective actions resulting from a home energy audit, including floor and wall insulation, attic airsealing and new attic insulation. The home is now very comfortable in all seasons. I am now replacing the roof with standing seam metal and have to make a decision re: attic venting. As far as I can tell, there are no current soffit vents anywhere, but there are a number of seeming randomly placed eyebrow vents, a solar fan, and numerous other types of vents. My issue is that the unconditioned attic is very hot, even on mild days. I'd like to use that area for storage, but I think the temps would likely damage anything I put up there. As far as I know, I have no other current issues with moisture, and ice dams are clearly not an issue. So, my questions are:

    1) Should I add venting when I install the new roof? Current discussion with the roofer includes a ridge vent and multiple eyebrow vents near the lower roofline.
    2) If not, what would you suggest for reducing temps in the attic?

    thank you!


    1. GBA Editor
      Martin Holladay | | #82

      Your attic will never be a good place to store items that might be damaged by heat -- not unless you want to install insulation that follows the sloping roof, which would be an expensive proposition. Store your temperature-sensitive items elsewhere.

      Yes, you should include a ridge vent when you re-roof. You should also try to install some type of vent that is relatively low -- either gable vents (if you have gables) or vented drip edge flashing. (Here's a link to show you what vented drip edge flashing looks like: Link to vented drip edge .) Eyebrow vents are also possible, but they are more expensive and more aesthetically intrusive. All of that said, venting won't do much to lower your attic temperature, and there may be no need for the lower vents if your attic has no signs of moisture problems.

      If you really want to reduce your attic temperature, convert your vented unconditioned attic into an unvented conditioned attic. Here's an article that explains the required steps: "Creating a Conditioned Attic."

      1. davehica | | #83

        Thanks Martin, you convinced me not to try and make the attic into useable storage. However, I am still struggling with what to do for the intakes. The roofer is recommending a ridge vent only because the way the gutters are configured would make a drip edge difficult to install. The also don't recommend installing multiple intakes low on the roofline due to the extra roof penetrations. Everything I've read however indicates that ridge-only venting could do more harm than good.

        So I think my question is should I allow them to install the ridge vent without addressing intake? Or exclude the ridge vent entirely? Kind of at a loss here because the roofer doesnt seem to think this is a huge deal but I would like to avoid making a critical error that could introduce new issues.

        1. GBA Editor
          Martin Holladay | | #84

          Unless your house is at a high altitude, I'm assume that your climate is relatively mild. If you don't see signs of mold or excess moisture in your attic, you are probably worrying unnecessarily. If I were you, I would let the roofer install the ridge vent, and keep an eye on the attic. I doubt you'll ever have moisture problems up there.

Log in or become a member to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |