Last updated: October 2026
Do cooling pillows work? They do cool your body, and they change how sleep feels — but a meta-analysis of nine randomised trials found no change in measured sleep: not sleep onset, not efficiency, not time in REM sleep.
Key takeaways
- A 2025 meta-analysis screened 3,454 articles, included 9 randomised controlled trials, and found no differences in sleep onset latency, sleep efficiency, sleep stage proportions, wake after sleep onset or total sleep time (Pasquier et al., Journal of Thermal Biology).
- The same review confirmed the cooling worked — lower core body temperature, p < 0.05. These products do what they claim physically, just not to measured sleep.
- Certainty of evidence was graded very low to low, with high heterogeneity: a weak evidence base, not a settled verdict.
- A 2025 crossover trial in 34 adults found a large gain in perceived sleep quality (p = 0.001, d = 0.92) and no significant change in any objective measure; its authors wrote that this “may in part reflect expectancy, or placebo effects” (Stevenson et al., Clocks & Sleep).
- What did move measured sleep was humidity and air movement: at the same 26 °C room temperature, radiant cooling produced higher sleep efficiency and shorter sleep latency than convection cooling (Park et al., Scientific Reports).
- In the one experiment manipulating temperature directly, mild skin warming shortened sleep onset by 26%, about 3.09 minutes, while core temperature changes were ineffective (Raymann et al., American Journal of Physiology).
- A pillow meta-analysis of 9 high-quality studies in 555 participants found pillow design did not influence sleep quality (SMD 0.047, p = 0.703) but did reduce neck pain and waking symptoms (Pang et al., Clinical Biomechanics).
Do cooling pillows work for measured sleep?
Not according to the best pooled evidence available. In 2025, Florane Pasquier and colleagues published a systematic review with meta-analysis in the Journal of Thermal Biology, asking whether body cooling produced by bedding — mattresses, toppers or pillows — improves night-time sleep in healthy adults aged 18 to 64. They searched PubMed, ScienceDirect and Web of Science to 4 March 2024, screened 3,454 articles, and found 9 randomised controlled trials meeting the criteria.
The pooled result was flat: no differences between cooling bedding and the habitual sleep environment on sleep onset latency, sleep efficiency, the proportion of the night spent in each sleep stage (N1, N2, N3 and REM), wake after sleep onset, or total sleep time. The review was pre-registered (PROSPERO CRD42021275369) and graded its own certainty of evidence as very low to low, noting high heterogeneity.
That caveat cuts both ways. A very-low-certainty null result is not proof that cooling bedding does nothing; the trials were few, small and inconsistent enough that a real effect could still be hiding. What it rules out is confidence in the opposite direction — and confidence is exactly what the product pages sell.
What did the cooling actually change, then?
Your body temperature. The same review reported that its included studies did show an effect on body cooling, with lower core body temperature (p < 0.05) on an acute basis. The physics is not in dispute: a surface engineered to pull heat away from you pulls heat away from you. The gap is between that measurement and the sleep measurement.
| Trial | Design | What improved | What did not |
|---|---|---|---|
| Pasquier 2025, meta-analysis of 9 RCTs | Cooling bedding vs habitual environment | Core body temperature (p < 0.05) | Sleep onset, efficiency, stages, wake after sleep onset, total sleep time |
| Stevenson 2025, 34 adults, 14 nights | Randomised crossover, cooling on 7 nights / off 7 nights | Perceived sleep quality (p = 0.001, d = 0.92), thermal comfort | Actigraphy metrics, heart rate, heart rate variability, respiratory rate (all p ≥ 0.05) |
| Park 2025, 7 women | Randomised crossover, radiant vs convection cooling at 26 °C | Sleep efficiency, sleep latency, “refreshness” (radiant) | — (both conditions held the same air temperature) |
Why does a cooling pillow still feel like it works?
Because the feeling is real even when the recording is not. The clearest demonstration comes from Shauna Stevenson and colleagues in Clocks & Sleep, October 2025. Thirty-four healthy adults (20 women, 14 men, aged 30 ± 5) used a temperature-controlled mattress cover for 14 nights in a randomised, counterbalanced crossover design — the temperature feature on for seven nights, off for seven. Sleep was measured by wrist actigraphy, with heart rate, heart rate variability and respiratory rate captured by sensors in the cover.
Every daily perceived outcome favoured the cooling condition (all p < 0.05), and perceived sleep quality improved substantially: p = 0.001, effect size d = 0.92. Objective sleep and biometric measures showed no significant differences at all (all p ≥ 0.05). The authors named the likely reason themselves, writing that the perceptual benefits “may in part reflect expectancy, or placebo effects.”
This is the same pattern we found in the pink noise research. If you wake up believing you slept well, that belief is part of your day — but it is not the same claim as “you slept longer”, and it is the second claim that bedding pages tend to make.
Does body temperature matter for sleep at all?
It does, and the detail is stranger than the marketing. Sleep onset normally follows a decline in core body temperature, driven by increased blood flow to the skin and heat loss. The question is which part of that sequence you can push from outside.
In 2005, Roy Raymann, Dick Swaab and Eus Van Someren tested exactly that in the American Journal of Physiology, collecting 144 sleep-onset latencies from eight healthy subjects while directly manipulating core and skin temperatures within the comfortable range. Warming the proximal skin — the trunk, not the hands and feet — by only 0.78 ± 0.03 °C around a mean of 35.13 °C changed sleep-onset latency by 26%, or 3.09 minutes (95% CI 1.91 to 4.28) against a mean of 11.85 minutes. Sleep came faster when the skin was warmed, even though subjective comfort dipped slightly.
Manipulating core temperature (0.20 °C) and distal skin temperature (0.74 °C) was ineffective. The one lever that causally shortened sleep onset was mild skin warming — close to the opposite of what a cool-touch surface is built to do. “Colder is better” is not a fair summary of this literature.
4 things that separate the trials that worked from the ones that did not
- Humidity, not just temperature. Radiant cooling produced lower relative humidity at the same air temperature — and it was radiant that showed higher sleep efficiency and shorter latency.
- Air movement. Radiant cooling also produced less air flow. Two rooms at 26 °C are not the same room if one moves damp air across you.
- What is being measured. Perceived sleep quality and actigraphy sleep efficiency are different outcomes, and cooling bedding has a far better record with the first.
- Where the heat sits. A pillow cools a small area under your head while a temperature difference lasts; room and bedding systems act on the whole surface all night.
What moved measured sleep: the bed microclimate
The most useful 2025 result is also the least intuitive. Insung Park and colleagues, in Scientific Reports, ran a randomised crossover trial with room air fixed at 26 °C, comparing two ways of getting there in seven healthy women: convection cooling, which supplies cooled airflow, and radiant cooling, which cools through chilled surfaces while a reheated refrigeration cycle keeps condensation from forming. Subjective sleep was assessed with the Oguri-Shirakawa-Azumi Sleep Inventory, objective sleep with EEG.
Radiant cooling produced lower relative humidity and lower air flow. “Refreshness” scored higher — and on the objective side, the part the cooling-bedding meta-analysis could not find, sleep efficiency was higher and sleep latency was shorter. The authors read this as the cooling and dehumidifying method changing thermal comfort despite a similar ambient temperature, and that change affecting both subjective and objective sleep.
Seven participants, all women, in a controlled setting makes this a signal rather than a verdict. But it points somewhere specific: what moved EEG sleep was how heat and moisture left the bed, not how cold the surface felt to the hand. If you sleep hot, ventilating and dehumidifying the bed may be a better-aimed intervention than a cool-touch cover, and it costs a fan.
Do pillows change sleep quality in general?
Mostly no, which is the context that makes the cooling result less surprising. Johnson Pang, Sharon Tsang and Allan Fu reviewed the pillow literature for Clinical Biomechanics in 2021, searching six databases to September 2020. Thirty-five articles met the inclusion criteria, of which 9 were high-quality studies involving 555 participants.
Their meta-analysis found that pillow design did not influence sleep quality in people with chronic neck pain (SMD 0.047, p = 0.703). What pillows did change was pain and waking symptoms: rubber pillows reduced neck pain (SMD −0.263, p < 0.001), waking pain fell (SMD −0.228, p < 0.001), neck disability fell (SMD −0.506, p = 0.020) and satisfaction rose sharply (SMD 1.144, p < 0.001). Their interpretation adds the detail that matters most when choosing one: cervical alignment in side-lying may be driven not by the filling material but by the shape and loft of the pillow.
Those are specifications you can check before buying — the FirmNest Extra Firm Contour Pillow, for example, is 60 × 40 cm with a 12–10 cm contour profile. The realistic expectation for any pillow, cooling or not, is comfort and morning neck symptoms rather than a longer night, which is the frame we use for the best firm pillow, side sleepers and neck pain.
Who should still consider a cooling pillow?
Someone whose main complaint is discomfort rather than sleep duration. If your pillow holds heat until you flip it at 2 a.m., a breathable or cool-touch cover addresses the thing actually bothering you, and the evidence for comfort and perceived sleep quality is the strongest part of this literature.
Someone chasing an extra hour, or more deep sleep on a tracker, is likely to be disappointed — and if you read sleep stages off a wrist device, our piece on what those REM numbers rest on is worth a look first. The levers with better evidence are a consistent schedule, light exposure and the bed microclimate, where our better sleep guide and the article on circadian rhythms begin. Persistent poor sleep is a conversation for a doctor, not a bedding purchase.
Frequently asked questions
Do cooling pillows work for hot sleepers?
For the heat, yes — cooling bedding lowered core body temperature with statistical significance in the trials pooled by Pasquier and colleagues. For sleep itself, the same review found no change in onset, efficiency, stages, wake after sleep onset or total sleep time, at very low to low certainty. If your complaint is feeling hot, that is the complaint the evidence supports addressing.
Can a cooling pillow make you sleep too long?
No trial has reported oversleeping as an effect of cooling bedding. The nine randomised trials in the 2025 meta-analysis found no change in total sleep time in either direction. A single long night after a bedding change is more plausibly explained by sleep debt or an irregular schedule than by the pillow.
Is a cooler bedroom better than a cooling pillow?
The one 2025 trial that moved EEG sleep did it by changing humidity and air movement at a fixed 26 °C room temperature, not by chilling a surface. That suggests ventilation and dehumidifying are the better-aimed levers. A pillow cools a small area for a limited time; the room acts on the whole night.
Does a colder surface help you fall asleep faster?
Not in the one experiment that tested temperature directly. Raymann and colleagues shortened sleep onset by 26% — about 3.09 minutes — by mildly warming the proximal skin, while core and distal skin manipulations had no effect. Warm feet and a cool head is a reasonable reading of that work; “colder is better” is not.
Why do cooling pillows get such good reviews?
Because perceived sleep quality genuinely improves. In the Stevenson crossover trial every daily perceived measure favoured the cooling condition, with an effect size of d = 0.92 for perceived sleep quality, while no objective measure changed. The authors attributed part of that to expectancy or placebo effects. Reviews capture the experience, which is the outcome that did improve.
Does memory foam make sleeping hot worse?
Dense foam holds heat longer than open, airy fills, which is why ventilated and shredded constructions exist. Firmness and loft also shift with temperature in viscoelastic foam, so a cold room can make the same pillow feel firmer. If your pillow feels different from how it did in the shop, our guide on break-in versus the wrong loft separates the two.
How long should I test a change like this?
The trials used short run-ins for a reason. Stevenson and colleagues gave participants at least three familiarisation nights before seven nights per condition. A week per condition, with your wake time held steady, is a sensible minimum before judging anything — and the evidence-ranked fall-asleep methods are a cheaper first experiment than new bedding.
Choosing on the specifications that the trials did move
The pattern across these five papers is consistent: temperature interventions change how sleep feels, while shape and loft change neck and waking symptoms. Neither reliably buys more measured sleep. That makes the sensible way to choose a pillow unglamorous — match loft and firmness to your sleeping position, keep the bed ventilated, and treat a cool-touch cover as comfort rather than as a sleep intervention.
Our own pillows are built around that first variable. Figures below are based on the maker’s production spec; the research above is published work, not an in-house trial. Readers with ongoing neck pain or persistent poor sleep should speak to a clinician before changing anything.
Advertisement
FirmNest Extra Firm Pillow
64 × 40 × 14 cm, 1.5 kg extra-firm memory-foam core.
See the FirmNest Extra Firm Pillow on AmazonProducts mentioned in this guide
Aeris Contour Pillow — 60 × 40 cm, 12–10 cm contour profile, 1.4 kg memory foam core
FirmNest Extra Firm Contour Pillow — 60 × 40 cm, 12–10 cm contour profile, extra-firm 1.4 kg core
