Cold house, wet floor.

Primary study → Read the LSU/USDA twelve-house raised-floor study (PDF)

Terms used here

• Relative humidity (RH): how full the air is of water vapor compared with the most it could hold at its current temperature. RH rises when air cools, even if no water is added.

• Dew point: the temperature at which the air’s water vapor starts to condense. It tracks how much water is actually in the air. Any surface colder than the dew point gets wet.

• Moisture content (MC): the weight of water in a piece of wood as a percentage of its dry weight, read with a moisture meter.

• Equilibrium moisture content (EMC): the MC that wood drifts toward when it sits long enough in air at a given temperature and RH.

• Permeance: how easily water vapor passes through a material. Vinyl and many polyurethane films are low-perm; they slow drying.

• Cupping: a board’s edges sit higher than its center, across the width of the board.

Numbers and formulas are collected in “The arithmetic” near the end. You can skip that section; the argument is complete without it.

What cupping is telling you

The National Wood Flooring Association describes cupping as edges high and center low, caused by “a moisture differential within individual pieces of flooring, usually excessive moisture on the underside.”² A floorboard gets wider as it gets wetter. When the bottom face is wetter than the top, the bottom grows wider and the edges lift.

In raised houses in south Louisiana this has a season. LSU AgCenter housing specialist Claudette Reichel describes it directly: “The wet subfloor wets the bottom of wood flooring, which causes the bottom of the planks to swell, causing cupping in summer.”¹ So the first diagnostic question is not what went wrong with the wood. It is why the underside of the floor is wetter than the top.

One floor, two climates

Above the floor, an air-conditioned room at 72°F and about 55% RH has a dew point near 55°F. In the LSU/USDA study, most houses averaged within a few points of 55% RH indoors through the summer.³ Wood in that air settles around 10% MC.⁵

Below the floor, the air is close to outdoor air. For an open crawlspace, NAHB notes that air temperature and dew point “differ only slightly from outdoors.”⁴ In New Orleans in June through August 2025, the outdoor dew point averaged about 76°F.⁸ Damp soil under the house adds water vapor on top of that.¹

Water vapor moves from wetter air toward drier air. In summer, that means upward through the floor, toward the air-conditioned room. The subfloor, and the underside of the hardwood on top of it, sit in the middle of that flow.^1,3

Why a colder thermostat makes it worse

The twelve-house study identified two effects that both run in the same direction.³

1. A bigger push. A colder house holds a lower indoor vapor pressure, which widens the gap between outside and inside. The study calls this a “larger inward vapor drive.”

2. A colder subfloor. In the study’s words, “for a given vapor pressure at the subfloor, a colder subfloor experiences a higher RH and therefore higher moisture content.” If the subfloor drops below the dew point of the crawlspace air, water condenses on it.¹

Insulation matters here. LSU notes that typical fiberglass batts “make it worse because the subfloor stays cooler.”¹ The batt insulates the subfloor from the warm crawlspace air, so the subfloor runs closer to the room temperature above, while humid air still reaches it through the batt.

The pattern in Figure 3 is the thermostat effect in measured form. Under vinyl, fiberglass-insulated subfloors averaged 18.0% MC in the house held near 70°F and 12.8% in the house near 78°F. Under carpet, the same comparison ran 14.1% to 11.8–12.4%.³ As PHP’s summary of the study reports, every house held at 78°F or warmer in summer stayed in a safe moisture range regardless of the insulation system.³ Read that summary.

Why the subfloor matters for the hardwood: NWFA’s installation standard wants the wood subfloor no more than 4 points of MC above solid strip flooring (under 3″ wide), and no more than 2 points above solid plank (3″ and wider).⁷ A subfloor that climbs through the summer pulls the bottom of the boards along with it.

The finish is a lid

The second half of the story is on top of the floor. LSU identifies “impermeable flooring, such as vinyl, laminate or a typical polyurethane finish on wood” as the condition under which cupping occurs and the subfloor problem worsens: “the subfloor just gets wetter and wetter all summer.”¹ The twelve-house study measured the same direction. Subfloors were generally wetter under vinyl tile than under carpet, and the thermostat effect was “more pronounced” under vinyl.³

Finishes do not stop moisture outright. The USDA Wood Handbook says coatings retard, but do not prevent, moisture change in wood.⁵ The practical point is that a low-perm layer on top slows drying toward the room while the crawlspace keeps adding moisture from below. LSU lists solid hardwood (not engineered) with a high-perm water-based polyurethane among the more forgiving choices.¹ How much each layer matters, and what happens when low-perm layers stack above and below the subfloor, is the subject of Part 3.

What about the nightly humidity swing?

A common explanation is that daily humidity swings are wearing out the floor. The Wood Handbook is plain on this point: short-term daily fluctuations “tend to influence only the wood surface.”⁵ RH climbs at night mostly because the air cools, not because water is added; PHP covered that in The dew point doesn’t drop with the thermometer. Summer cupping over a crawlspace is driven by a moisture difference through the board that builds over weeks, not overnight.

What to do this summer, in order

1. Measure before you change anything. Write down the thermostat setting. Put a hygrometer in the room and, if you can safely reach it, one in the crawlspace. Take moisture readings of the flooring from above and the subfloor from below. NWFA says the moisture source for a cupped floor can be verified “only with a moisture meter that takes readings of the underlying subfloor.”² Field judgment: read a flat room too, so you have a comparison.

2. Raise the summer setpoint toward 76–78°F. That is LSU AgCenter’s guidance for raised houses.¹ If the house feels muggy at 78°F, the cooling system is not controlling humidity well, which is a reason to fix humidity control rather than to run the house colder. NAHB recommends designing HVAC to control humidity at higher setpoints such as 78°F.⁴

3. Take water away from under the house. Rainwater should not run under the house. A plastic ground cover limits soil moisture, but LSU warns that water must not pool on top of it.¹ Look for dryer vents, A/C condensate lines, and plumbing leaks discharging below the floor. PHP’s dryer case shows how much one source can add.

4. Do not add fans blowing summer air under the house. LSU: “Bringing more humid air to cold subfloor materials can actually worsen the problem in humid weather.”¹ The numbers are in Crawlspace ventilation by the numbers.

5. Do not sand the floor yet. Cupped boards usually flatten as they dry, which NWFA says “may take weeks, months, or even an entire heating season.” Sanding a cupped floor while it is still wet can leave it crowned once it dries.^1,2 Part 2 covers the waiting period and how to tell when it is over.

6. If it comes back every summer, look at the floor assembly. The twelve-house study found that foil-faced polyiso board and closed-cell spray foam kept subfloors in a safe range across the conditions tested. LSU says those systems let you use any flooring and any thermostat setting, and cautions never to insulate while the subfloor and flooring are wet.^1,3 Part 3 takes up the layers in detail.

When it probably is not the thermostat

• Cupping in one spot. If it is near a dishwasher, refrigerator ice maker, exterior door, or bathroom, look for a local leak first. NWFA lists each of these as a cupping source.²

• Cupping that never eases. If the boards stay cupped through a full heating season, NWFA says they are probably permanently deformed, or the moisture source is still active.²

• Boards lifting off the subfloor. That is buckling, a more serious condition with its own causes (Part 2).²

• A sealed, conditioned crawlspace. The two-climate picture does not apply in the same way. NAHB notes that floor-covering permeability does not matter when the space below is conditioned.⁴ Look for a failed dehumidifier, a water leak, or a gap in the enclosure.

The arithmetic

Optional. The argument above stands without it.

What this article does not claim

• The twelve-house study monitored 12 occupied houses for one year and measured subfloor MC, not hardwood MC. Some data points rest on few observations.³

• A 76–78°F setpoint is a strong lever, not a guarantee. Water sources, drainage, and the floor assembly still matter.

• Open, vented, and sealed crawlspaces behave differently. This article is about open and vented crawlspaces under air-conditioned houses.

• This is a diagnostic explainer, not a repair specification.

Sources

Related on PHP

• One year, twelve houses, six ways to insulate a subfloor

• The physics were always going to win (open-cell foam under sheet vinyl)

• Everybody blames the foam. The dryer was venting into the crawlspace.

• Crawlspace ventilation by the numbers

• The dew point doesn’t drop with the thermometer

Next in the series: Part 2, Cupping, crowning, buckling, gapping: what the shape is telling you.