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

Installing Rigid Foam Above a Concrete Slab

You can sandwich rigid foam between an existing concrete slab and new plywood or OSB subflooring

If your basement slab lacks sub-slab foam, you'll have to install the insulation above the slab. There are several ways to install subflooring above the rigid foam: you can use either one or two layers of plywood or OSB, and the subflooring can either be fastened to the concrete with long Tapcon screws or it can simply float above the foam. If you choose to have a floating subfloor, it's best to install two layers and screw them together.

If you’re building a new cold-climate home with a basement, you’ll probably want to install a continuous horizontal layer of rigid foam under the basement slab. Even though the rigid foam won’t save enough energy to justify its cost, it’s worth installing for another reason: it will reduce summertime condensation and mold accumulation, and will therefore help your basement smell better. (To learn more about sub-slab foam, see “All About Basements.”)

What if you’re living in an older house that has an uninsulated basement slab? If you want the benefits of floor insulation, you’ll have to install the rigid foam on top of your existing slab.

Three ways that slabs get wet

Before we discuss slab insulation, it’s worth reviewing the three basic mechanisms that can make a basement slab damp:

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. fwsolar | | #1


    You've apparently read my mind, once again, by publishing an article on a project already on the docket for my 40-year-old house. Or maybe it's just that my house has the usual suite of "normal" issues. Either way, thank you!

    I have a question regarding your mention of using interior poly for the basement floor in an interior-foam retrofit: should the poly extend up the walls and, if so, how far? In my situation, in would certainly be possible to run the poly up the walls as high as the foam will go, which is to the top of the concrete.

    I plan on using spray foam to insulate the top surface of the concrete wall and around the rim joist (between floor joist bays). That will be after air-sealing the rim joist area to the best of my ability. The floor joists rest on top of the concrete - they are not embedded.

    Any comments or thoughts are greatly appreciated!

  2. GBA Editor
    Martin Holladay | | #2

    You wrote, "I plan on using spray foam to insulate the top surface of the concrete wall."

    I'm not sure what you mean by "the top surface." There are two uncertainties:

    1. Are you using open-cell spray foam or closed-cell spray foam?

    2. Are you only insulating the top few feet of the wall, or are you insulating the entire wall, from the slab up to the top of the wall?

    If you are insulating the wall on the interior, from the slab all the way to the top, with closed-cell spray foam, you don't need any polyethylene. The closed-cell spray foam will be a vapor barrier, a water barrier, an air barrier, and insulation.

    For more information, see "How to Insulate a Basement Wall."

    1. fwsolar | | #3

      My thought is to use rigid foam (EPS) for both the walls and the floor, and running from the slab all the way up to the top of the walls.

      By "top surface" I mean the exposed horizontal surface of the wall, upon which the joists rest. It's here that I plan to used closed-cell spray foam, which will also run continuously to cover the rim joist etc.

      I will do the rigid foam work myself and let a local pro handle the spray foam.

      1. GBA Editor
        Martin Holladay | | #4

        In my experience, a layer of EPS foam provides a perfectly adequate vapor retarder in this location, so you don't need polyethylene.

        However, polyethylene is a vapor barrier, while EPS is only a vapor retarder. So if you want to be absolutely sure that there isn't any water vapor transmission from the concrete to the interior, you can install poly between the concrete and the EPS if you want. I would call it optional.

        Remember, installing polyethylene on the wall makes it hard to use adhesive to attach the EPS to the concrete wall.

  3. Reid Baldwin | | #5

    I am considering replacing basement flooring at my parent's house, which currently has carpet over concrete with no foam. Would I need to do anything special around existing partition walls?

    1. GBA Editor
      Martin Holladay | | #7

      Q. "I am considering replacing basement flooring at my parent's house, which currently has carpet over concrete with no foam. Would I need to do anything special around existing partition walls?"

      A. What type of flooring will you be installing? Will you install carpeting again? If so, most experts advise that it's better to have an insulated slab under the carpeting. Then again, if they've had carpeting for years with no complaints, their basement may be dry enough (or warm enough) to avoid problems.

      I'm not sure I understand you question about the partitions. Clearly, no one is advising you to install carpeting under the partition bottom plates -- so you'll work around the partitions.

      1. bwjames | | #25

        Honestly I think I'd be mildly concerned about the bottom plates on the partition walls. Current state, the carpet may be permeable enough to prevent moisture problems. Future state, If you suddenly have a material with low-perms over all of the slab except the bottom plates, you will see an increase in moisture at those plates. This could accelerate rot and mold problems in those partition walls. I believe this to be a fair concern.

        1. GBA Editor
          Martin Holladay | | #26

          When Reid posted his question, I wasn't sure what his concern was. It's possible that you have correctly identified his concern.

          Bottom plates installed directly on a concrete slab are supposed to be pressure-treated. If this partition has a pressure-treated bottom plate, as it should, I wouldn't worry.

          1. bwjames | | #27

            Hi, I'm Brad and I think I figured how to update my username!

            I agree that if the bottom plates are pressure treated and there is a capillary break, Reid will be fine. Pressure treated on its own will not do much of anything to prevent excess moisture from getting into the wall though.

            Reid's parents may also be lucky and just have a dry basement. Of course, there are two kinds of basements - those that are wet and those that will be wet!

  4. lukasmpeter | | #6

    Thank you for the very timely and excellent article!
    Do you have any advice on how to proceed if the concrete slab is sloped? I am planning to convert a garage into an ADU (accessory dwelling unit). I hope to avoid the use of self leveling concrete for environmental reasons. Would a leveling of dry and clean sand, on top of the EPS and below the plywood, work?
    I was thinking of placing pressure treated and ripped sleepers at 8 feet on center on top of the eps to create guides to level the sand. These sleepers would align with the short edges of the first plywood layer placed on top. The two layers of plywood and sand would be floating on top of the insulation.

    1. GBA Editor
      Martin Holladay | | #8

      My first reaction to your question was to suggest self-leveling compound, until you wrote that you are trying to avoid it. I'm not sure why you don't want to use it.

      I'd be leery of trying your idea (sand on top of the EPS). I'd be more likely to use crushed stone (carefully leveled) under the EPS. Perhaps other GBA readers have more suggestions.

      1. lukasmpeter | | #12

        Thank you Martin. The sand idea comes from a remodel experience in a very old multistory building where I found sand under the wood plank flooring between the support beams. I suspect it was installed to provide sound isolation btw the floors.
        I am trying to use less concrete to reduce the carbon footprint where possible. I do recognize that this may not always be possible.

    2. maine_tyler | | #15

      I'm in a similar situation.

      Would a foam subtractive method be totally crazy i wonder? I.e. level some rails to ride a sled with some foam eating tool on it. I could only see this being feasible if there was a tool properly suited that could do it quickly enough and collect dust.

      edit: for slightly unlevel, not many inches of course. I also realized that it'd be quite a waste of foam. I suppose you could use some different layers of 1-2 " foam to build up prior where its the lowest to minimize eating tons of foam.

      1. GBA Editor
        Martin Holladay | | #16

        Q. "Would a foam subtractive method be totally crazy, I wonder?"

        A. In my opinion -- yes, it would. Use floor leveling compound or crushed stone.

  5. Expert Member
    Peter Engle | | #9

    Roofing insulation manufacturers can provide customized, sloped insulation boards. Both EPS and polyiso should be available. This will be virgin material as it is manufactured custom for the job, but it would work, and probably be significantly cheaper than self-leveling compound. An added bonus would be that the insulation value would increase as you get closer to the primary exterior wall.

    If you're looking for a concrete (pun intended) solution, self-leveling compound would still not be the choice. These products are very expensive, and you would need a lot of it to fill the typical 3"-6" slope of a garage floor. Lightweight (gypsum) concrete would be a much better solution, or even just concrete grout - that's concrete without the bigger aggregate. Either one can be ordered loose enough to easily rake and trowel flat. You would want it to taper to a minimum of 1" thickness at the thin edge.

    1. lukasmpeter | | #13

      Thank you Peter. I’ll price the tapered insulation and compare it to the concrete leveling per your suggestion. The amount of cement used for this type of non-structural concrete would be very low and therefore have a relatively low carbon footprint.

  6. Expert Member
    KOHTA UENO | | #10

    Nice column, Martin!

    Install one or two layers of tongue-and-groove OSB or plywood on top of the rigid foam. If you install just one layer, you’ll need to fasten the OSB or plywood to the concrete slab with long Tapcon screws that extend through the rigid foam.

    This is the method a friend and I did in his basement. We were very happy with the results--per your other correspondents, it results in a very solid-feeling floor, with minimal "bounce." If there are "bouncy" spots, you can just add another Tapcon.

    I concur with the recommendation for adding a countersink to the Tapcon hole. I've generally found that Tapcons do not "pull" the same way as wood screws, so the fastener will not countersink on its own into plywood or OSB. So countersinks are required to keep the fastener heads flush.

    In the example shown below, we used 1" XPS (due to head height restrictions) and taped seams. My friend omitted the polyethylene.

    If you want a utilitarian basement floor rather than a finished space, you might consider installing a cementitious tile underlayment panel (something like Durock) directly on top of the rigid foam insulation. This type of panel is heavy and flat enough to stay in place by gravity, without any fasteners, and can be used as “industrial-look” flooring.

    This is the method I used in my basement--trying to have dry tool storage and workshop space; it went together very quickly and easily. Taped the 2" XPS seams, and put the Durock on top--XPS foam recovered from a dumpster, yeah! I won't be roller skating around down there, so there's no need for a high durability material. I definitely recommend cement backer board (Durock, Wonderboard) over fiber cement tile backer (Hardie)--the latter is "stiff" enough that a point load will "curl" up the other part of the board. The "floppy" Durock conforms to uneven surfaces. One issue with this system is that the edges of Durock are not durable relative to impact. So if you have a condition where the edge is exposed (e.g., door threshold between insulated and uninsulated slab), it needs some protection. That's another item on my project list. :/

    FYI--the foam + cement board retrofit is what we used in the basement of the Utica NY retrofit you wrote about (

    1. GBA Editor
      Martin Holladay | | #17

      Thanks, Kohta. I appreciate the tips and the photos.

    2. chandler01950 | | #73

      Looks like what I'm going to do. In one of your photos it looks like you foamed these rigid foam sheets to the concrete. For air sealing I'm guessing since you didn't use poly. Makes sense but I figured I'd ask the the question. My "slab" is rough and spalling. I've vacuumed it clean many times and will do so again before I install 2" of rigid foam.

      1. GBA Editor
        Martin Holladay | | #74

        I'm not sure if Kohta will see your comment, so until he does, I'll add my 2 cents. You can use canned spray foam as an adhesive to hold the rigid foam to the slab if you want -- the key, of course, is applying just the right amount of foam (not too little, not too much) and keeping some kind of weight on top of the rigid foam until the spray foam has a chance to cure. Too much foam, and not enough weight, could allow bumps to form in the floor.

        Personally, I wouldn't depend on spray foam to air seal the seams of the rigid foam, unless you are talking about installing the canned spray foam along the rigid foam seams after the rigid foam has been installed. I would use tape at the seams for air sealing.

        1. Expert Member
          KOHTA UENO | | #75

          Thanks for commenting Martin--I saw it in "recent comments."

          To directly answer the question--I just loose-laid the foam sheets down on the floor, no foam adhesive. But using some foam (in the right amounts, per Martin) would do a nice job of filling any air gaps below the foam, and helping (possibly) deal with an uneven substrate.

          I used housewrap-type tape to seal the foam joints as an air barrier.

  7. Expert Member
    KOHTA UENO | | #11

    One more thing--I'd recommend EPS and XPS over polyisocyanurate in this on-slab application. This is admittedly old (BETT 1982) data, but it indicates that polyiso loses a lot more of its R-value due to moisture uptake than XPS or EPS (in buried conditions). Of course, there are those other studies out there that says EPS gains much less moisture than XPS.... paid for by the EPS industry. Of course, the EPS density matters as well.

  8. Expert Member
    Peter Engle | | #14


    While I agree that polyiso might not be the best choice for this application, I wouldn't really call these conditions anything like underground with soil contact. Unless your basement is constantly wet, the foam is going to see dry conditions nearly all the time. I would think the moisture uptake (and R-value reduction) would be small.

    Seems like a good idea for grad student work.

  9. Expert Member
    KOHTA UENO | | #18

    Hi Peter. My thought is that in the Northeast, there are lots of older housing stock with poorly poured basement slabs. I often refer to them as borderline dirty concrete/conrete-y dirt. :) So of course, no vapor control underneath. That was the slab condition at the two projects that were the source of the photos above.

    Therefore, the conditions at the slab-to-foam interface are likely to rise to 100% RH for most of the year--so it is an area where you want a material that is resistant to moisture uptake. I agree that this is not as challenging as burying foam in dirt, but I still would recommend EPS/XPS over polyiso here.

  10. Expert Member
    Peter Engle | | #19

    Point taken. The wetter the conditions, the less optimal polyiso performance is going to be. I just wanted to point out that it shouldn't be ruled out in all cases, because there are plenty of slabs that are dry, but uninsulated where polyiso might be a good choice.

  11. user-6733106 | | #20

    I've got a split level home built in 1979 in upstate NY (zone 5). Kind of funny, at the time, they built the mid-level section with a vented crawlspace, with common interior doors for access to the crawl space. Combine that with fiberglass insulation between the floor joists, and spotty rigid insulation installed below the floor joists in the crawl space and no insulation on the foundation walls or floor... it's not at all ideal. I'm looking to seal up the vents in the crawlspace, add some air movement from the conditioned space into the crawl space and add rigid insulation to the walls and floors. Are there any issues leaving the rigid foam exposed on the walls and floors of the crawlspace area?

    Assuming I utilize Dow XPS insulation, I believe for crawl spaces, I can install without a thermal barrier or ignition barrier. The link provides an ICC-ES Evaluation Report:

    1. GBA Editor
      Martin Holladay | | #22

      Q. "Are there any issues leaving the rigid foam exposed on the walls and floors of the crawlspace area?"

      A. As far as I know, the code requirements for covering rigid foam with an ignition barrier or a thermal barrier are the same as the code requirements for these barriers over spray foam. For more information on the issue, see "Thermal Barriers and Ignition Barriers for Spray Foam."

      A thermal barrier is more stringent than an ignition barrier. Covering foam in a crawl space with the less stringent barrier (an ignition barrier) is only possible if it meets the following criteria: if the crawl space is only accessed for repairs or maintenance; if there is no easy access to the space; and if it isn’t used for storage.

      As far as I know, XPS needs to be covered with either a thermal barrier or an ignition barrier for fire safety. But if you have any doubts, you should contact your local building department to find out how local inspectors interpret the code in your jurisdiction.

      1. user-6733106 | | #24

        appreciate the reference to the other article - very helpful.

  12. StephenHood | | #21

    Hi Martin,

    Thanks for the timely article. I'm currently building a home in North Louisiana (Climate Zone 3). We have a slab on grade foundation (post tension stressed). However, for a room of the house we are using as a home theater, we dug down 2 feet. So, it's not a full basement...just a partial sunken room. The intention is to build a staggered seating platform to have 2 rows of seating. So, the rear portion of the room will be built up on two layers of 2x12s to bring it level with the rest of the 1st floor in the house. However, the front portion of the room (about half) will remain at grade level in the room (2 feet lower than the rest of the first floor). We have installed a French drain around the base of the exterior walls on this room and applied water sealing compound and dimpled mat. On the inside of the walls, I have applied a layer of 3/4" EPS foam. Under the concrete slab, we used a 10 mil sheet of poly.

    Now, my question is what should I do on the floor? We were planning to install a 1/2" rubber underlayment (similar to horse mat material) glued to the concrete. On top of that we would glue a layer of 3/4" Advantech paneling (this is a sound isolating technique prescribed by the home theater designer). Then, we would install carpet on top of the Advantech (with the typical carpet pad) in the front half of the room. In the rear half, we would build the stage/riser platform on top of the Advantech.

    However, reading your article, I'm wondering if I should put some EPS foam down on the concrete before putting in the rubber mat underlayment. And would I really need 2" of EPS being in Zone 3 where it is predominantly warm...but can be very rainy/wet? Could I get away with something like 1/2" or 3/4"? Or should I just stick with the original plan and use a special glue that will also serve as a water barrier?

    Thanks in advance!

  13. GBA Editor
    Martin Holladay | | #23

    In your climate zone, you probably don't need rigid foam for this floor. It's a judgment call. For more information on this issue, see "Determining Sub-Slab Rigid Foam Thickness."

  14. craigcarter | | #28

    I have a project that includes converting a storage room in a garage to a conditioned laundry room space. The entry from the house to this area drops about 5 - 6 inches. The client would like the new laundry room floor to be level with the existing house floor. There would be two exterior doors that would need to be installed as well - one into the garage and another that leads to the outside. Initially, I was advised to have poly, 2x4 sleepers, rip down 2x's for joist nailed to the sleepers, install rigid foam between the joist and install 3/4 subfloor to achieve the desired height...this seems like a lot of work when compared to this articles assembly. I thought about this assembly as it is similar to an article I recently read on a "slabless slab" installation that appears to be very similar. If I follow the assembly instructions and not fasten the subfloor into the concrete slab are there any potential concerns in this area with the use of concrete tile as the finished floor? Do I need to leave a gap around the perimeter of the rigid foam and framing or should this be snug?

    1. GBA Editor
      Martin Holladay | | #35

      Q. "If I follow the assembly instructions and I do not fasten the subfloor into the concrete slab, are there any potential concerns in this area with the use of concrete tile as the finished floor?"

      A. You should contact the tile manufacturer to determine the manufacturer's subflooring requirements. If you have two layers of 1/2-inch or 3/4-inch plywood or OSB screwed together, however, I imagine you'll be fine.

      Q. "Do I need to leave a gap around the perimeter of the rigid foam and framing or should this be snug?"

      A. A little gap is a good idea -- but you should caulk the gap.

  15. MarkAno | | #29

    Martin/Dana/GBA Community:

    Regarding Dana’s “month long Polk Party” post when he was talking about rigid foam compression concerns, I have the following question:

    Dana references EPS Type-II. Is EPS Type I (1lb density) acceptable/strong enough for above basement slab use?

    My proposed Spec is 2” of Type I (1 lb) EPS (10psi) for the following basement floor assembly:
    (Starting from the basement concrete slab going up):
    Concrete Slab
    2” EPS
    6 mil Poly
    2 layers of 1/2” Advantech OSB
    1 layer 3/4” Advantech OSB
    Luxury Vinyl Plank

    - No load bearing walls will be placed on assembly.
    - Typical basement furniture, bed, Home Theater chairs

    Heavier loads may include:
    - home gym equipment (treadmill, free weights on stand)
    - kitchen / wetbar with granite countertops
    - pool table

    1. GBA Editor
      Martin Holladay | | #36

      Q. "Dana references EPS Type-II. Is EPS Type I (1lb density) acceptable/strong enough for above basement slab use?"

      A. This is a complicated topic. For a deep dive into the issue, I suggest that you read this article: "Foam Under Footings."

  16. Deleted | | #30


  17. DavidLeff | | #31

    I'd rather do the floating floor than fasten through to the concrete. In this situation, why can't one use a single 3/4" layer of high quality OSB/plywood and glue the OSB/plywood to the rigid insulation underneath it, rather than use two thinner layers of OSB/plywood screwed to each other?

    1. Expert Member
      MALCOLM TAYLOR | | #32


      That's good question, and I can't think why not, although I've never heard of anyone doing it. Tacking the foam to the slab with adhesive would probably help, and you would need a lot of weights to keep the plywood flat while the glue dried, but it should work.

      Can anyone think of why it isn't a good idea?

    2. bwjames | | #33

      I did a little testing with this exact theory 3/4" over 1.5" of 15psi foam, there was a little flex movement at the joint under foot from compression. Two layers of 1/2" ply with staggered seams comparatively was rock solid.

      That said, was the deflection any more than a 3/4" ply joint between floor joists? Not sure but that would be an interesting comparison.

  18. DavidLeff | | #34

    My reasoning is:
    - It should be at least as good as the Amdry product that has a layer of foam and OSB glued together. It should be better than the Amdry product because of the ability to tape and stagger the seams.
    - High quality 3/4" OSB weighs approximately 75 lbs per 4'x8' sheet - which is about the same weight as the equivalent area of Durock (mentioned above).

    Bradley, did you try tongue and groove? Was it plywood or OSB?

    1. nbatalla | | #83

      Hi David,

      5 years later and I'm facing a similar situation. Like you, I want to use 3/4" osb and float it over the foam, but have concerns with deflection. What did you end up doing? Thanks.

  19. bradpj53 | | #37

    I have an unusual application for this approach. My home is earth sheltered concrete and steel, massively built. It was an experimental 1980's build that was wonderfully over-engineered by the previous owner, a retire d NASA engineer; 12" filled and rebar-ed block walls on 4' deep 3x3 footings. The original earthen rooftop (over a poured concrete slab-on-steel, supported by steel trusses) had to be enclosed due to failure to be sufficiently water resistant. So I now have a second floor space that never existed originally. I am in the process of removing approximately 50 tons of earth from this space (24 x 34 feet).

    This new space, with its 10 foot ceiling, will be my new workshop. Ultimately it will be fully climate controlled, but that is months away. Insulation was provided by 6" of EPS under the dirt, but that is going along with the dirt. My needs are to soundproof the floor because of machinery noise, to protect the living space beneath, and to provide at least some degree of insulation both to keep out the summer heat from the downstairs living space until project completion, and to provide future flexibility in HVAC for the two separate spaces.

    My plan had been to lay down rubber acoustic mat and subfloor, and insulate ASAP, but now with the decision to add insulation I am wondering about 2" of Iso, or even possibly Roxul 80, with a double 1/2" or even double 3/4" floating subfloor over that. The main difference from a Rec rood situation is that my workshop equipment is relatively heavy (some vintage pieces are cast iron and > 1,000#) and the locations are not fixed; I have found that workshop layout requires flexibility of machine placement as need changes over time. Will this approach over-stress the foam (OR Roxul) and what are my other options? Thanks!


  20. GBA Editor
    Martin Holladay | | #38

    I'm not sure what a "Rec rood situation" is -- maybe you meant to write "rec room"?

    Briefly, you question is above my pay grade. Lots of issues here. My instinct tells me that you shouldn't put 1,000-pound cast-iron power tools on top of rigid foam. You need to get help from a structural engineer and a sound engineer.

    I'm also concerned with the continuity of your home's barriers -- the continuity of the insulation layer (now that you are planning to peel away some or all of the insulation from your home's roof) and especially the continuity of your home's waterproofing layer (now that you are introducing a new seam between the earth-sheltered portion of your home and the new second story). Good luck with that.

  21. kevin_in_denver | | #39


    You forgot the fourth way that slabs get wet - plumbing failures.
    As the owner of over one hundred rentals houses over the years, I wouldn't put anything organic like plywood on any basement floor. You'll be tearing all it out a few weeks after the water heater fails.
    I would use high density XPF under a luxury vinyl tile.

    1. NeilMac | | #70

      I was wondering if 47 PSI XPS might be good to go with just LVT overtop, and no plywood. Any insight on that?

  22. joe_eugene | | #40


    I'm constructing a log house on a floating slab in southwest Wisconsin, and am making plans for a wood floor on top of the slab.

    I'm nearly certain the contractors did not install a vapor barrier beneath the slab. They did put in rigid foam (2 inches) on the sides of the slab, and beneath the slab, although they neglected to put it under the 12-inch perimeter where the logs sit. The foundation has no drains, pipes or holes in it.

    For my floor, my plan is to put down a layer of rigid foam over the slab. On top of that I'll put 1x12 kiln dried tongue and groove pine boards for my subfloor. And then I'll put a wood floor over that.

    Two questions:
    1) Should I put down that vapor barrier you mentioned on top of the slab, before I lay down the rigid foam?

    2) Will my pine boards (already purchased) be able to float like the OSB boards without causing issues?


  23. GBA Editor
    Martin Holladay | | #41

    It's astonishing that any concrete contractor could be so ignorant that he or she would forget to install polyethylene under the slab. If there is no rigid foam under the logs (!), the slab could be damp directly under the logs -- which is certainly undesirable. So make sure that you install a capillary break of some kind between the slab and your logs.

    Arguably, the 6 mil poly above the slab won't be necessary if you plan to install rigid foam. That said, some types of rigid foam are somewhat vapor-permeable, so it makes sense to install poly above the slab. Poly is cheap.

    I don't think that pine boards will make a good subfloor here, since you aren't able to nail the boards to joists. Use tongue-and-groove plywood or tongue-and-groove OSB.

  24. joe_eugene | | #42

    Thanks, Martin.

    Yeah, I was pretty stressed when I saw them in the middle of pouring concrete with no vapor barrier. We considered ripping the whole thing out, but that didn't feel right. So beneath the logs, I put sill seal directly on the slab, and then I put two full two-inch-by-ten-inch black locust boards (so four inches thick) between the sill seal and logs. Black locust is incredibly rot resistant, so I feel pretty good. (They say if you put a black locust post in the ground, the hole will wear out before the post!)

    I won't be doing OSB for the subfloor, as I already bought the pine boards (and I already am not using other pine boards I bought for the ceiling after you'd wisely recommended drywall instead). I just can't afford it. So maybe I should do a 6 mil poly sheet, then 2x4 sleepers (I'm thinking of using cedar as it's somewhat rot resistant), then rigid foam between the sleepers, then my tongue and groove pine over those. Would that work?


  25. user-7604598 | | #43

    We are doing a small partial garage conversion and installed rigid foam above the concrete slab with two staggered half-inch plywood layers on top. To bring the floor level with the house, we used 4.5" of rigid foam. We have finally gotten the building permit guy to stop asking for floor framing but now he wants a manufacturer's statement that this is an approved application for the rigid insulation and that it will meet the 40 pounds per square foot pressure that the code requires.. I called the manufacturer's rep who, of course, couldn't say this is an approved application. He confirmed that the product we used will withstand 20 psi (the charts I found online only said 16 psi but I think either would be sufficient to meet code). All of the data sheets say that this product is for vertical applications. I found one that said it was also for "floor assemblies" but I don't know if that will suffice. Any suggestions on where I can find an official statement that would satisfy my building codes officer?

    1. GBA Editor
      Martin Holladay | | #44

      I don't have an "official statement" I can point to. But you should probably check out this article: "Foam Under Footings." That article includes lots of technical information and links to references discussing the bearing capacity of various types of rigid foam.

      Of course, any GBA readers who have encountered skeptical building officials should chime in with any helpful advice.

  26. user-7604598 | | #45

    Thanks Martin. I looked at that article but one of the links on bearing capacity (the one I thought would have good data) wasn't working. I am really out of my element with this stuff but I can load the permit office up with pdfs full of indecipherable technical specs. It is worth a try and much cheaper than hiring an engineer for my 10' x 10' garage conversion.

    1. GBA Editor
      Martin Holladay | | #46

      If you are talking about the link to the document titled "EPS Geofoam Data Sheet," there was indeed a problem with the link. I just corrected the problem. Here is the corrected link: "EPS Geofoam Data Sheet."

  27. user-7604598 | | #47

    Thank you! That was exactly what I need.

  28. bmac89 | | #48

    After installing platon and 2" of EPS, I am considering installing a regular 1/2" plywood for first course, then screwing, and possibly glueing, a 3/4 T/G course over it. Mostly this is due to lack of availability of 1/2" t/g plywood. The specifications above suggest two 1/2" layers of t/g plywood. Is it ok to do one course of regular plywood or do both layers need to be t/g.

    I know its an option to use one course and screw into the slab, but I prefer not to do that. Also, prefer not to use OSB, particularly on the first layer since it will need to hold the screws on the second layer above.


    1. Kevin_in_PA | | #51


      I'm considering the same approach. How were the results for you?


      1. bmac89 | | #62

        Not finished yet with overall project (and just noticed your questions) Took some time away, but the floors have been down for over a year and built a bunch of walls on top and everything seems pretty solid. No odors. The 3/4' T/G was hard to get down solo, but with rubber mallet and a block of wood, was able to get it in. An assistant would have been a real help.

  29. GBA Editor
    Martin Holladay | | #49

    Your plan sounds fine to me.

    1. bmac89 | | #50

      Thank you

  30. mjshuster1 | | #52

    Hey Martin- Hoping you might be able to advise on a specific use case.

    We're converting a garage to a music space and would like to put down laminate plank flooring. What we have right now is concrete slab on grade and we're concerned about heat loss (we're in Massachusetts). I don't believe there's insulation or vapor barrier under the slab, and although we haven't done a moisture test yet, there doesn't seem to be any serious moisture problems.

    The tricky part here is we only have about 1.5" of total height to work with for all the layers including insulation/subfloor + floor, so I'm trying to maximize the thermal r-value within that. I've been looking at all-in-one products like Dricore, but I'm not sure it's the best bang for the buck within the 1.5". Another idea is to lay a dimple membrane like DMX/Delta-FL, then something like a 1" rigid EPS/XPS onto that. I was thinking that approach would give us a nice vapor barrier & air gap as well as decent R-value.

    If we're trying to do a floating floor, would you recommend one approach over the other? And if the rigid foam route, would it be possible to accomplish that without having to glue or secure w screws?

    Any advice would be greatly appreciated, thank you!


    1. GBA Editor
      Martin Holladay | | #53

      I don't recommend the use of DriCore. For more information, see "Fixing a Wet Basement." Note especially Comment #6 on that page by Mike Guertin.

      1. mjshuster1 | | #54

        Thanks Martin- even for slab on grade? What if we did a 6mil vapor barrier sheet down first, then the dricore?

        For the rigid foam route, would it be sound to place a sealed layer of dimpled membrane directly on the slab, then 1" XPS (Foamular 150?), then the laminate planks just floating right on top of that? I believe those 3 layers alone would be close to the 1.5"of height I have to work with. Probably can go over that by a 1/4" but not more.

        Thanks again! -Matt

        1. GBA Editor
          Martin Holladay | | #55

          My advice on installing rigid foam above a slab can be found in the article on this page. Your question about installing laminate flooring is a question for the laminate flooring manufacturer -- flooring manufacturers get to decide what types of substrates are suitable, so contact the flooring manufacturer for installation instructions. (Note that these instructions are often available online.)

  31. lauraje | | #56

    Hey martin,
    I'm planning on refinishing my basement in BC (zone 6). It is a concrete floor with concrete walls up to above grade. When demolishing the existing finishes (with the aim to better waterproof the concrete from the interior) I discovered mould on the concrete walls and floor. The floor was carpet installed over thin poly (likely 3-4mil) and the walls were 2x4 fibreglass batt insulated with the same thin poly over them. I assume that the poly is what trapped the moisture and created the mould issue, but there seems to be conflicting info in the article above and those referenced, as to whether or not you should use a vapour barrier on the floors and walls. My plan is to paint the walls (and floor?) with drylok or equivalent, insulate the floor with 1" eps, two layers of 1/2" ply or osb, insulate the walls with 1.5" eps then frame a 2x4 wall in front (and on top of the plywood,) insulate that with mineral wool batts, and drywall/carpet for the finishes... Do you see any issues or missed steps with this plan? Budget and ceiling heights both play into this.


    1. GBA Editor
      Martin Holladay | | #57

      Your floor developed mold because the first condensing surface was cold. Installing 1 inch of EPS should solve the problem.

      Your walls developed mold because they were insulated with fiberglass batts (a type of insulation that is air-permeable) directly against the concrete, allowing that the warm, moist interior air to contact the cold concrete walls, where condensation occurred. Installing 1.5 inch of EPS (which is an air barrier) against the concrete walls will solve this problem. (Make sure that you install the EPS in an airtight manner.) So your plan will work.

      1. lauraje | | #58

        Thanks Martin for your help.

        I am also wondering if i was to go with a click together vinyl plank floor in place of the carpet would this create a vapour barrier over the subfloor assembly? I ask because I am now looking to create a laundry space in the basement and hoped to use the same subfloor system.


  32. mrknobber | | #59

    Hello! I have a low ceiling basement in Toronto (zone 4) with 3/4 of the wall below grade. I'm following your guidelines here for the floor 6mil poly sheet and 1" rigid insulation sealed with good euro tape. I'd prefer 2 layers of wood to prevent me having to penetrate everything to the concrete. What is available to me is 7/16" OSB or 3/8" ply (ply a small price increase). Out of those two which would you recommend? I hear lots of 'never osb' on the interwebs, with a smattering of 'osb is fine' but I also will loose 1/8" using osb over the ply with 2 layers. Will the thinner ply layers be enough? The foam board is 20psi.

    Lastly, I left the poly a bit longer and it goes up the wall a small bit. Where is the best place to trim it? I thought of taping it on top of the wood subfloor and putting the studs over top. I'll be having spray foam installed to 3" thick on the outside walls if that makes a difference.

    I have so many questions but those are the two that are bothering me. Thanks so much!

  33. GBA Editor
    Martin Holladay | | #60

    Mr. Knobber,
    I would probably prefer two layers of 3/8-inch plywood to 2 layers of 7/16-inch OSB. But the two layers of 3/8-inch plywood are still a little thin. My guess is that the plywood will probably work, although other GBA readers are invited to chime in.

    Where you trim the polyethylene doesn't matter much. Taping it to the wood subfloor sounds fine to me.

    1. mrknobber | | #61

      Thank you Mr. Holladay. We are going with 1/2" ply, as we read between the lines, ha ha.


  34. StephenRenee | | #63

    I have a similar situation to Laura -- Zone 7a, Gulf Islands BC. Insulating an above(ish)-grade freezewall-and-slab in a 1970s house. The freezewall is 8" thick and 1-1.5ft high, but 6-7" of that is above the slab. Nice for headspace, bad for insulating. The MAJOR source of heat-loss on FLIR is this freezewall, and insulating it on the outside isn't feasible as most of it is poured on sandstone or surrounded by concrete patios.

    My tentative plan is to fur the perimeter walls out to 2x8 (well, essentially frame a second wall out, really) using up the extra 4" on the freezewall. That'll sort the top out, I hope, and give lots of room for Rs. Then mulling over something like a sandwich of 2" flash-and-bat rigid EPS + spray foam in the cavities; remainder Roxul. Then some version of the insulated subfloor discussed here: moisture or dimple barrier, 2" of foam, 3/4" ply sandwich.

    But... I'm completely stuck on the subfloor-freezewall-wall transition detail.

    What could work here? I considered cutting some 1 or 1.5" foam strips to sit on the subfloor's insulation, then carry it above the sill plate... cover that with baseboard, maybe... ? Not sure technically OR aesthetically how to handle this, and I can't find much online, either. I've attached a photo of the lip for reference. Any thoughts?

    PS: The staining is from an ancient gutter problem, long fixed -- no mould at all behind any of the fibreglass bats here or elsewhere, though I did find some mouse poop. That... just adds to the R-value right?! ;)

  35. GBA Editor
    Martin Holladay | | #64

    Your description is confusing, so I'll try to understand your photo so I can provide advice.

    The photo shows a gray floor at the very bottom. Is that a concrete slab or some other material?

    Next (moving from the bottom up) I see a wooden chase of some kind. What's that? 1x4s or 1x6s making some kind of box? What is under the lumber?

    Next I see a 2x4 stud wall. Is this the wall you want to make thicker (transforming it into a wall that is equivalent in thickness to a 2x8 wall)? Or did I misunderstand?

    1. StephenRenee | | #65

      Hi Martin, thanks and sorry if I mixed up any terminology here! It's kind of an unusual setup.

      I didn't want to call it a "stem wall" -- it isn't built quite like that, but that's the closest idea. Imagine a stem wall where the slab was poured half a foot below the top of it (done to gain headroom, I think). The stem wall is 8" thick, so the framing sits on the back half of it. I've seen this called a "freezewall" but that may be wrong.

      Taking the photo from bottom to top:

      1. The grey floor is concrete slab with a thin industrial carpet on top.
      2. The wood is 1/2" trim pieces glued to the side and the top of the protruding stem wall; plan is to pry those off it.
      3. Yes, that 2x4 stud wall is what I want to make thicker, increasing the total equivalent thickness to 2x8. That will cover all but 0.5" of the stem wall's top edge.

      Adding a subfloor reduces the stem wall's height in the room from 6" to 3".

      But that leaves me with a ~0.5" wide x ~3" high concrete lip to figure out how to insulate and tie into the new wall and the new floor.

    2. StephenRenee | | #66

      Alright: a picture's worth a thousand words, but a drawing is worth a million... ! This may help. Proposed insulation additions on the right, including where I'm stuck...

      1. Expert Member
        Akos | | #67

        Set the bottom plate of the interior wall extension an a piece of higher density rigid. Since this is a non-load bearing wall addition, there is no issues with sitting on foam. Now the stem wall is completely wrapped in foam and there are no thermal bridges.

        If you want to simplify your transitions and baseboards, a better option might be to make an even thicker wall so that the inside finished drywall surface is in-line with the inside of the foam over the concrete. You can now cover the foam plus bottom of drywall with a piece of tall baseboard.

        Your initial pictures show a fair bit of water damage in the existing wall, those need to be fixed before you do anything.

      2. GBA Editor
        Martin Holladay | | #68

        Akos is right. Align your new wall framing to ensure that your interior finish material (presumably, drywall or paperless drywall) is co-planar from ceiling to floor. You want the upper wall framing to be situated so that the finish materials are continuous. There is no harm in a thicker wall -- it provides room for more insulation.

        1. StephenRenee | | #69

          Thank you both for the advice, that is an elegant solution that I've been blocked on for awhile now. Really appreciate it! A bit less space isn't ideal, but space that's toastier, more livable, and with no energy cost for most of the year sure is a decent trade-off.

          Akos: yep, definitely; a gutter let out there for years, but has since been dealt with. A thorough assessment is on the list. Though an initial aggressive poking found sound wood, I'm sure some remediation will be needed here or there.

  36. seanpatrickrice | | #71


    Would there be any concern with 2 layers of 1" XPS, seams taped on each layer, with the layers running perpendicular to each other, then the 2 layers of subfloor, installed perpendicular to each other and fastened together?

    Would two layers of 1" XPS create the potential for moisture on condensation between them? would one layer of 2" XPS, with seams taped be a better option?

    in the particular application I am talking about, I would be enclosing an open car port with raised slab-on grade. The exterior walls will be wood, double bottom plate sealed to foam gasket sealed to slab, and filled with fiberglass insulation. (or would there be a better wall assembly in this instance, for roughly the same cost?)

  37. GBA Editor
    Martin Holladay | | #72

    Q. "Would there be any concern with 2 layers of 1-inch XPS, seams taped on each layer, with the layers running perpendicular to each other, then the 2 layers of subfloor, installed perpendicular to each other and fastened together?"

    A. No performance concerns, although green builders try to avoid the use of XPS, which is manufactured with a blowing agent that has a high global warming potential. EPS or polyiso would be preferable.

    Q. "Would two layers of 1-inch XPS create the potential for moisture on condensation between them?"

    A. No.

    Q. "Would one layer of 2-inch XPS, with seams taped be a better option?"

    A. Not from a performance perspective.

    Q. "The exterior walls will be wood, double bottom plate sealed to foam gasket sealed to slab, and filled with fiberglass insulation (or would there be a better wall assembly in this instance, for roughly the same cost?)"

    A. Ordinary frame walls filled with fiberglass batts are inexpensive, but to perform well, such a wall needs meticulous air sealing and very careful installation of the batts to prevent voids or insulation compression.

  38. Avatar_Lucas | | #76


    I have a question. I am considering buying a stone house in Zone 4 near Philadelphia. Taken back to the bare walls and joists there is some discussion about both the basement walls, floor and walls above grade. At this stage there appears to be no insulation anywhere. The architect has called for simple batt and wooden stud insulation but no more on the walls and nothing for the basement floor.

    It seems to be agreed that closed cell foam above grade and up against the stone wall is preferable? This is also the suggestion you have for below grade but with a Vapor barrier? If I was to use a rigid insulation board for the floor and then the double plywood solution either gravity or screwed down with tapcons, how would the rigid insulation meet the closed cell foam insulation at the wall threshold?

    1. Avatar_Lucas | | #77

      Hi Martin, have you seen this question? Thank you

  39. user-7529177 | | #78

    I'm curious if anyone has had success eliminating the OSB/plywood subfloor by installing a floating floor material directly over EPS or XPS on top of the slab? If so, what material did you use? We are planning a basement retrofit and have ceiling height limitations.

    1. NeilMac | | #79

      I haven't done this yet, but I am hoping to. I spoke with a Rep at an EPS manufacturer, and they said it would be possible with a 40+ PSI foam. I think I will go 60 PSI just to be safe.

  40. sterilecuckoo58 | | #80

    I do wish I had found this column BEFORE executing the plan, which looks very much like the article's detail.

    Capillary break - The story I had learned involved providing a capillary break between the floor foam and the wall foam. To this end, I cut rolls of roofing vent material into 2-in. strips, and laid these flat on the 2-in. floor foam against the wall. (While we had not had any identifiable inbound moist, though while we had temporarily removed gutters, runoff was able to get in and make a mess before placing foam on the floor.) I used spray foam to fill gaps between the floor foam and the concrete wall. My first layer of 7/16" OSB butts to the vent material. My first layer of wall foam (2-in. EPS Type 1 - 1 lb/ft³ so maybe I have Type 2?) sets on the 1/2-in. vent material. The second layer of wall foam sets on the first layer of 7 /16" OSB.

    Air and water sealing - I used a bead of permeable sealant along the bottom of the 2nd layer of wall foam where it meets the first (lower) layer of OSB. Where foam joints were not tight, I also used spray foam, end either captured it with seam tape and first layer of OSB, or let it cure uncontrolled and cut to "grade". I used permeable seam tape on the floor foam seams, including those with spray foam filler. The wall foam was treated in a similar manner, except we sealed the 2nd layer of seams with tape. (Spray foam was used for voids). I sealed the OSB top layer with an acrylic based wood sealer stain (a bargain high quality permeable mis-tint). The OSB joints received a non-permeable self leveling sealant.

    To fasten or float - I found it necessary to fasten the first layer of OSB to the concrete floor. Tap Cons are indeed a pain. We glued and screwed the second layer to the first layer butted to the 2nd layer of wall foam. That was a lot of effort. The OSB is reclaimed Sheathing grade, not subfloor grade. It seemed to need more screws than it deserved in order to stay flat to the first layer. Perhaps this is on account of "reclaimed" or perhaps because it isn't rated subfloor.

    Two layers OSB or one for the floor - For the 2nd half of the basement I am using one layer of the nominal 3/4-in. T&G Advantech Subfoor. This too is screwed to the concrete floor. The 17 screws per sheet is feeling pretty good. I am doing 21 under equipment / appliances. I am liking this better than the two layers. Thicker is not feasible. I have room for a 7/16" floating plank floor and 1/4" leeway before bumping into the 7'-0" minimum ceiling height.

    Floating plank floor directly over foam - We tried this at the beginning. Even with 25 psi Type 9 EPS, there was insufficient rigidity at the joints. I also tried Amdry panels over the foam (I thought perhaps in lieu of 2 layers OSB or Plywood) and laid the plank floor over that. This too was not as solid as one might like (might have been had we screwed the Amdry through the foam to the concrete, but that seemed counterproductive). I'll try / floating the the Amdry for a moist garage floor with foam and the Advantech subfloor.

    Rather than eliminate the plywood / osb, one might consider eliminating the slab. Replace the slab with drainage aggregate, polyethylene, foam and then the engineered wood subfloor. The story told is that it works for new construction. (I am uncertain where the poly goes... maybe it goes over the foam)

    The basement no longer smells like a basement - The previously installed stud walls with FG kraft faced batts were indeed moldy. Some of the gypboard was also grody (which, I believe is derived from the slurring of gross and moldy). We have reused most of the lumber as furring on the exterior. We cleaned it and treated with an anti mold solution of washing soda (aka sodium carbonate), trisodium phosphate and sodium bicarbonate. We are also treating all new, reinstalled, repurposed and existing hidden lumber with a boric acid based solution.

    I have used 1x4's and 2x4s as furring for the foam on the wall and as a screw base for the 5/8-in. gypsum board. I figure to repurpose non-moldy fiberglass batts to fill the 1-1/2 in. void. These are mechanically anchored to the wall and captured at the top and bottom.

    Existing Untreated bearing-wall base plates - I have treated this one troublemaker with propylene glycol. I am considering improving penetration by drilling small holes between the studs. I'll insulate and air and water seal the assembly with provision to inspect / disassemble easily. This might also be a good place for a capillary break along its sides.

    Stairs - We have not yet modified the floor at the base of the stairs (indoor and bulkhead). Resetting the treads changes the head clearance. Yes, there's a fix for that too.

  41. GPR100 | | #81

    Hi Martin,

    First off, I've found the site and your articles to be immensely helpful while working through some plans for refinishing the basement rooms in my 45-year-old home.

    You mention in the article, "If you have a load-bearing wall in your basement, the bottom plate of that wall should rest on the concrete. (In most cases, it will be a pressure-treated plate.) But if you are talking about a non-load-bearing partition, it’s perfectly OK to install the bottom plate of the partition on top of your new plywood or OSB subfloor."

    The two bedrooms I'm redoing will get 6 mil poly > 1.5" XPS > 2 layers of OSB for the subfloor. The above applies directly because the partition/dividing wall between the two rooms (running through the center of the basement) is 2x6 framing (including an 'embedded' jack-post) that supports a load bearing header.

    My question is, what is the best practice for insulating the floors up to this center wall? I'm wondering if I should run the poly sheet up the load bearing wall a few inches or use some specific approach while re-finishing?

    The bottom plates on that center 2x6 wall are PT lumber. Still, I want to do my best to keep moisture out of that center wall between the rooms, given that the framing will remain in direct contact with the slab floor. (I should also note that this center wall runs all the way to the exterior slab basement wall on one side. So I'll be working around that when insulating and framing the exterior basement walls as well...)

    Thanks in advance for any advice!

  42. sterilecuckoo58 | | #82

    Like GPR100, I have a plate on the concrete slab. Unlike GPR100, ours is untreated. I don't want to encapsulate it with foam, but running an air barrier (like a WRB) up the wall, across the stud void and back down keeps the indoor air away from the plate. Adding mineral wool to the stud void under the WRB allows for drying and limits the thermal bridging to the studs. Adding insulation to the stud edges to a point 2 to 4 in. above the floor foam increases the R value of the stud-thermal bridge. I wonder at what point does this become overkill.

Log in or become a member to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |