Pillow Height Studies: What Research Says About Loft
No pillow height study names a single best height for everyone. An X-ray study of 16 young adults suggested 10 cm for back sleeping; a 2016 biomechanical model found taller pillows raised neck pressure; a side-sleeper study found individualized, shoulder-width-based heights of about 9.7–11.8 cm worked best. The pattern across the studies: height matters — and the right height depends on sleep position and body size.
Applies to: if you want the research behind loft charts — what pillow height studies actually measured, not what retailers claim.
Looking for the one pillow height study that settles the loft question? Here is the honest version. One X-ray study of 16 young adults suggested a 10 cm pillow for lying on the back. A 2016 pillow height study found taller pillows raised pressure on the neck. A side-sleeper pillow height study used shoulder width to predict height. None settles it — and if you want the research behind loft charts, the reason none of them can is the interesting part.
Three studies, three methods
The phrase "studies show" does a lot of heavy lifting in pillow marketing. But this corner of the literature is thin: each pillow height study is small and methodologically unusual — nobody has run a big randomized trial of pillow height and neck pain. What exists is a handful of laboratory investigations, each answering a narrower question than the marketing implies. Here they are, side by side, with exactly what each one measured.
The X-ray study: Kim and colleagues, 2015
The most cited pillow height study is a Korean hospital study that took lateral X-rays of 16 healthy adults aged 20 to 30 while they lay on their backs on pillows of 0 cm (flat), 10 cm and 20 cm
Kim, H. C., et al. (2015) — X-ray study, n=16 asymptomatic adults aged 20–30, supine, pillows at 0/10/20 cm; measured T1 slope, C2–7 Cobb angle, neck tilt. PMC4623167As pillow height rose, two spinal angles climbed: the T1 slope went from about 12.9 degrees at 0 cm to 17.3 degrees at 10 cm to 28.8 degrees at 20 cm, and the C2–7 Cobb angle (the curve of the neck itself) rose from 10.3 to 14.9 to 17.4 degrees. Neck tilt tended to fall. On that basis the authors recommended 10 cm as the height giving the most normal cervical lordosis — and then immediately hedged: 10 cm was "just a numerical value" that ignored comfort and satisfaction, and "not an absolute figure."
The computer model: Ren and colleagues, 2016
A second pillow height study never touched a real sleeper. It built a finite-element computer model of the head-neck complex and simulated four pillow heights from H0 to H3, tracking pressures and geometry (Ren et al., 2016). Raising the modeled height pushed the cervical angle up 66.4 percent, the lordosis distance up 25.1 percent, cut the kyphosis distance 43.4 percent, and significantly raised cranial and cervical pressures. This is where the "65% more pressure" framing comes from — a modeled number, not a measurement on people. The same paper noted that optimal height did not correlate with simple head-and-neck dimensions, which is a quiet argument against one-size-fits-all charts.
The side-sleeper study: Beihang University
The third pillar is a side-sleeper study from Beihang University that took the individualization idea seriously: it derived a shoulder-width-based index (called H-phi) to predict each person's optimal pillow height (Beihang University, n.d.). A medium individualized height of 9.74 to 11.76 cm with neck support gave the cervical curve closest to natural standing and the lowest musculoskeletal force. Notably, the highest comfort score went to a slightly taller sub-low range of 11.76 to 13.78 cm — comfort and geometry disagreed. One honesty note: the year and journal for this study were not captured in our records, so it is cited undated and flagged for verification before republication.
The supporting cast
Three more pillow height studies orbit these three. Yamada and colleagues (2023) tested a setup method holding the cervical spine at about 15 degrees of anterior tilt from the bed surface, adjusted in 5 mm increments; half of 84 participants reached a clinically meaningful drop in neck pain (Yamada et al., 2023). Jiao and colleagues (2025) ran a small EMG pilot (n=15) and found that a pillow height equal to 1.0 times shoulder width produced the lowest neck-and-shoulder muscle activity and the best comfort scores, both at p<0.001 (Jiao et al., 2025). And a systematic review of pillow effects from 1990 to 2019 concluded that heights between 5 and 11 cm maintain muscle-tension balance and cranio-cervical alignment (Airiti Library systematic review, 2021).
Scroll sideways on small screens.
| Study | Method | Who or what was tested | Key finding | The limit |
|---|---|---|---|---|
| Kim et al. 2015 — Korean J Spine | Lateral X-rays, supine | 16 healthy adults, 20–30, pillows at 0/10/20 cm | T1 slope 12.9→17.3→28.8°, C2–7 Cobb 10.3→14.9→17.4° as height rose; 10 cm suggested for normal lordosis | Young healthy sample; comfort never measured; authors' own caveat |
| Ren et al. 2016 — PeerJ | Finite-element computer model | Modeled head-neck complex, heights H0–H3 | Cervical angle +66.4%, pressures significantly raised at high loft | No real sleepers; no pain or sleep outcomes |
| Beihang — undated | Individualized design, shoulder-width index | Side sleepers | 9.74–11.76 cm with neck support: best curve, lowest force; 11.76–13.78 cm: best comfort | Year/journal not captured — flagged for verification |
| Yamada et al. 2023 | Clinical setup method | 84 participants, ~15° cervical tilt, 5 mm steps | 50% reached a clinically meaningful pain drop | Method study, not a height recommendation |
| Jiao et al. 2025 | EMG pilot | 15 participants | 1.0× shoulder width: lowest muscle activity, best comfort (p<0.001) | Small pilot; short-term muscle signals only |
| Airiti review 2021 | Systematic review, 1990–2019 | Pillow literature across three decades | 5–11 cm window supports alignment and muscle balance | A range, not a number — and mostly lab measures |
These studies cannot be ranked by "best number" because they measured different things in different bodies: X-ray angles in healthy twenty-somethings lying awake on their backs, pressures in a computer, comfort and muscle signals in side sleepers. A number from one method does not transfer to another. That is also why the oft-quoted "10 cm" and "65% more pressure" live in different universes — one is a suggested height for a specific sample, the other a modeled pressure change. Treat each pillow height study as a separate finding, which is what they are.
What they agree on
Different methods, different samples, different decades — and still the findings converge. Here is what the pillow height studies agree on, stated as plainly as the data allows.
Agreement 1: height moves the geometry — measurably, every time
No pillow height study has ever found that height doesn't matter. The X-ray study moved spinal angles with every centimeter; the computer model moved pressures and curvature; the EMG pilot moved muscle activity. Whatever else is uncertain, loft is a real mechanical variable — the neck's position changes when the pillow's height changes. If you take one thing from this page, take that: "pillow height" is not marketing vocabulary. It is the single best-measured thing about pillows. Every pillow height study we reviewed agrees on that much, even when they disagree on the number.
Agreement 2: fixed numbers lose to individualized ones
The strongest through-line in the research is that one fixed number keeps losing to personalization. The Beihang side-sleeper study's headline result is that an individualized height — 9.74 to 11.76 cm, derived from each person's shoulder width — beat every fixed alternative on cervical curve and musculoskeletal force. Jiao's EMG pilot found the same pattern in miniature: the height that minimized neck-muscle activity wasn't a chart number, it was 1.0 times the individual's shoulder width. Even the 2021 systematic review of 35 pillow trials (on pain, not height) concluded that shape and height matter more than material (Pang, Tsang & Fu, 2021) — and height, in practice, is always relative to a body. That is the through-line of every serious pillow height study in this set. The surprise buried in this agreement: the research keeps pointing away from "the right height" and toward "the right height for you."
Agreement 3: comfort and geometry do not always agree
This is the agreement nobody's marketing mentions. In the Beihang study, the height range with the best cervical curve (9.74–11.76 cm) was not the height range with the best comfort score (11.76–13.78 cm). The X-ray study never measured comfort at all — the authors said so themselves. So a pillow can be geometrically "correct" and still feel wrong, and a pillow can feel wonderful while holding the neck in a worse position. Any height advice that ignores this split is telling half the story.
Agreement 4: no study links a specific height to long-term pain
Here is the honest gap. The studies measure angles, pressures, muscle signals and comfort — none of them followed sleepers for months to see which height produced less neck pain. The 2021 systematic review found that pillow interventions reduced chronic neck pain (rubber pillows SMD −0.263, p<0.001), waking pain and disability, but it pooled pillow types, not heights (Pang, Tsang & Fu, 2021). So "this height relieves pain" is a claim the literature has never directly tested. Anyone who tells you otherwise is selling, not citing.
The pattern the studies actually support
Loft is mechanically real; individualization beats fixed numbers; comfort and geometry can disagree; and the height-to-pain link is untested. That is the whole honest state of pillow height research — everything else on this page is detail.
What none of them can tell you
⚠ Limits of this evidence
- Comfort was barely measured. The X-ray study never asked anyone how the pillow felt; the Beihang study found comfort and geometry pointed at different heights.
- The samples are narrow. Sixteen healthy twenty-somethings lying awake on their backs is not a population — it is a pilot. Nobody has tested heights in older adults or people with existing neck pain.
- A computer model is not a sleeper. The 2016 model's pressure numbers are simulations; they describe what the physics predicts, not what anyone felt.
- One position at a time. Back-sleeper findings do not transfer to side sleepers, and no pillow height study exists for combination sleepers who change position overnight.
- No pain outcomes. Angles, pressures and muscle signals are proxies. Whether a given height means less neck pain months later is the question nobody has run.
The marketing version
"A pillow height study proved 10 cm is the scientifically correct height — so this 10 cm pillow is clinically validated."
What the study actually said
The authors called 10 cm "just a numerical value" that ignored comfort and satisfaction — "not an absolute figure" — in 16 healthy young adults, measured awake, on their backs. Validation, it is not.
There is a broader point here about how to read pillow height studies at all. Each pillow height study is a laboratory answer to a laboratory question: what happens to this angle, this pressure, this muscle, at this height, in these people, right now. The questions buyers actually ask — will this height stop my morning stiffness, is it worth switching, how long should I try it — sit one or two inferential steps beyond what was measured. The honest studies say so in their discussion sections. The marketing skips the discussion sections. Reading any pillow height study this way — methods first, numbers second — is the skill this page is trying to teach.
First: does sleeping at your "geometrically ideal" height actually change how your neck feels in the morning? Plausible, untested. Second: how long does it take to know whether a new height works — days, weeks? Also untested; the studies measured single sessions. Until someone runs a months-long randomized trial of assigned heights with pain diaries, every height recommendation is an educated inference from short-term lab measures, not a tested treatment. We state this plainly because it is the single most important sentence on this page. Hold every future pillow height study you meet to the same bar.
Does pillow height affect neck alignment?
The phrase "does pillow height affect neck alignment" sounds technical, but the idea is simple: your pillow props your head at some height, and your neck bends to accommodate it. The studies just gave that bend numbers. Here is what each term means, translated.
The three angles, in plain language
T1 slope is the tilt of the top of your upper back where the neck begins — think of it as the launch angle of your neck. C2–7 Cobb angle is the curve of the neck itself (the lordosis, the gentle backward C). Neck tilt describes how the whole assembly leans. In the X-ray study, raising the pillow from flat to 20 cm pushed the T1 slope from 12.9 to 28.8 degrees and the neck curve from 10.3 to 17.4 degrees, while neck tilt fell (Kim et al., 2015). In plain terms: a taller pillow bends the neck more sharply and tilts the whole column forward. That is the core mechanical finding of every pillow height study that measured angles. The paper's discussion noted that a T1 slope between roughly 13 and 25 degrees tends to accompany better sagittal balance, and the 10 cm pillow landed inside that window — which is the actual reasoning behind the 10 cm suggestion, not a round-number preference.
The computer-model pillow height study tells the same story in pressure language: at high modeled loft, the cervical angle jumped 66.4 percent and cranial and cervical pressures rose significantly (Ren et al., 2016). And the EMG pilot adds the muscle view: the wrong height makes the neck and shoulder muscles work harder all night, with 1.0 times shoulder width minimizing that effort (Jiao et al., 2025).
Text equivalent: the diagram shows a side sleeper with the head on a pillow, and a straight line running from the neck through the shoulder — the "level" position the pillow height studies treat as their reference. Arrows show the two failure modes: the neck bent upward (pillow too tall) and sagging downward (pillow too flat). This is the single visual check behind the mirror self-checks the research literature describes: ear, shoulder and spine in one line.
Is 10 cm the best pillow height?
"Is 10 cm the best pillow height" is the most searched question in pillow height research, and the answer is a lesson in how findings get flattened. The 10 cm figure comes from one pillow height study with very specific conditions, and every one of those conditions limits how far the number travels.
Who 10 cm was tested on — and who you might be
| The study's conditions | What that means for you |
|---|---|
| 16 healthy adults, aged 20–30 | If you are older, or have existing neck pain, you were not represented. |
| Lying on their backs, awake, X-rayed | Side and stomach sleepers were not tested; sleep itself was not tested. |
| Only three heights: 0, 10 and 20 cm | Everything between — 5 cm, 8 cm, 12 cm — is unmeasured in this study. |
| Comfort and satisfaction not measured | The authors said so explicitly; the "best" height might feel wrong. |
| No pain outcomes tracked | 10 cm optimized an X-ray angle, not anyone's morning stiffness. |
| Preliminary study, authors call for larger work | The paper describes itself as a starting point, not a verdict. |
The fair reading: within its narrow conditions, 10 cm is the best-supported single number in the literature — it is also the only single number any pillow height study has ever recommended, which is not the same as being right for you. The authors' own framing is the one to keep: 10 cm is "just a numerical value," "not an absolute figure." Retailers quoting it as a clinical fact are quoting half the sentence. No later pillow height study has overturned the 10 cm finding, but none has generalized it either — the sample remains 16 people.
Where 10 cm sits against the other numbers
Placed beside the rest of the evidence, 10 cm looks less like an answer and more like a data point. The Airiti systematic review's supported window is 5 to 11 cm — 10 cm sits near its top edge. The Beihang individualized range for side sleepers is 9.74 to 11.76 cm — 10 cm sits at its bottom edge, and that study adds neck support as a co-variable the X-ray study never considered. Yamada's clinical method sidesteps height entirely and targets a 15-degree cervical tilt adjusted in 5 mm steps. Three different "right answers," each conditional. The honest synthesis is not "10 cm wins" — it is "10 cm is one well-documented point inside a wider, person-dependent range." This is why the phrase "a pillow height study found" should always be followed by "in whom, and how."
The shoulder-width method for side sleepers
The shoulder-width idea is disarmingly simple: when you lie on your side, the gap your pillow must fill is the distance from the mattress to your neck — and that distance is roughly your shoulder width. Two pillow height studies tested this logic from opposite directions.
The Beihang side-sleeper study formalized it as an index (H-phi) computed from shoulder width, then tested heights around each person's predicted value (Beihang University, n.d.). The individualized medium height — 9.74 to 11.76 cm with neck support — produced the cervical curve closest to natural standing and the lowest musculoskeletal force. The twist, again, is comfort: the highest comfort scores went to a slightly taller 11.76 to 13.78 cm band. Geometry and comfort disagreed by about two centimeters.
Jiao's 2025 pilot came at the same question through muscle electricity: 15 participants slept at heights scaled to their own shoulder width while EMG recorded the sternocleidomastoid and trapezius muscles (Jiao et al., 2025). The height of 1.0 times shoulder width produced the lowest muscle activity (p<0.001) and was rated the most comfortable (p<0.001). Small sample, short-term signals — but it is the only pillow height study that tested the shoulder-width rule directly rather than deriving it, which makes it the most practically useful of the set.
What to measure yourself
The research version of the self-check is straightforward, and it is the closest thing this literature offers to a personal answer. No pillow height study prescribes these exact steps — they are our synthesis of the methods the studies themselves used:
- Measure your shoulder width — the horizontal distance from the outer edge of one shoulder to the other, or simply the gap from mattress to neck when you lie on your side.
- Start near that number — the EMG pilot's 1.0-times rule, or the Beihang 9.7–11.8 cm band for average adults.
- Mirror-check the line — lie in your usual position and have someone check (or photograph) whether your neck sits level with your spine, not bent up or sagging.
- Weight comfort equally — the studies show geometry and comfort can disagree by centimeters; a height that aches is the wrong height regardless of the angle.
For the full step-by-step version with illustrations, see how do you check your pillow height at home? And if you sleep on your side and wonder about pillow width as well as height: how wide should a side sleeper's pillow be?
🔬 The loft lab: drag the slider, read the studies
Move the slider from 0 to 12 cm. The readouts below show what the actual pillow height studies found at — or near — that loft. Loft: 10 cm of 12
Back-sleeper lens · Korean X-ray study, 2015
10 cm: the study's suggested height — T1 slope 17.3° and neck curve 14.9°, inside the balance window. The authors' caveat stands: comfort and satisfaction were never measured.
Side-sleeper lens · Beihang individualized study
10 cm: inside the individualized 9.74–11.76 cm band that gave the best cervical curve with neck support.
What loft charts claim — and what they cannot prove
Search for pillow height advice and you will find confident charts: side sleepers 4–6 inches, back sleepers 3–5 inches, stomach sleepers under 2–3 inches or none. This consensus — widely repeated, including in AI-generated overviews — is the "loft consensus" retailers operate on. Here is the problem: none of the guides we reviewed cite a study for these ranges. They are convention dressed as science. A pillow height study can anchor a number; a chart that cites none is just a chart.
That does not make them useless. Compare the convention with the research: the back-sleeper band of 3–5 inches (about 7.6–12.7 cm) brackets the X-ray study's 10 cm suggestion. The side-sleeper band of 4–6 inches (about 10–15 cm) overlaps the Beihang 9.7–11.8 cm curve-optimal range at its lower end and the 11.8–13.8 cm comfort-optimal band above it. The convention and the research rhyme — but rhyming is not citing, and a chart that cannot name its source cannot be checked. For the loft concept itself, see what is pillow loft, exactly?
The one research-backed range
If you want a range with a citation attached, the Airiti systematic review (1990–2019) is the closest the literature offers: pillow heights between 5 and 11 cm maintain muscle-tension balance, cranio-cervical alignment and pressure distribution (Airiti Library, 2021). It is a wide window — six centimeters wide — which is itself the point: the research supports a range and individualization, not a number. Retail charts that quote half-inch precision are implying a certainty the studies do not contain.
A chart needs numbers to look useful, and "about 10 cm, adjust to comfort" does not look like expertise. So charts slice the continuum into neat bands — 4–6 inches, 3–5 inches — that suggest measurement where there was only tradition. The studies, by contrast, keep discovering that the answer is "it depends": on position, on shoulder width, on whether you weight geometry or comfort. Precision in the charts is a design choice; uncertainty in the studies is a finding. Trust the finding.
Can my pillow cause neck pain?
"Can my pillow cause neck pain" is the worry behind most loft searches, and it deserves a direct answer: no pillow height study has ever tested whether a particular height causes pain. The studies measure what height does to angles, pressures and muscles — not whether those changes produce pain over time. That is the gap, stated plainly. Until a pillow height study follows sleepers for months with pain diaries, the causation question stays open.
What the research does say is adjacent and useful. The 2021 systematic review of 35 pillow trials found that changing pillows reduced chronic neck pain (rubber pillows: standardized mean difference −0.263, p<0.001), waking pain and neck disability — and that shape and height mattered more than fill material (Pang, Tsang & Fu, 2021). So pillows as interventions can ease existing pain, and height is one of the levers that matters. For the full trial picture, see do cervical pillows work in trials?
The practical reading: if you wake with neck pain or stiffness most mornings, your pillow's height is a reasonable suspect — not a convicted one. The studies support experimenting with height (the shoulder-width method above is the most evidence-anchored way to do it) while watching symptoms over a couple of weeks, rather than treating any number as a diagnosis. And if the stiffness is the main problem rather than the height question, start with why do you wake up with a stiff neck every morning?
Questions people actually ask
Effect of pillow height on the biomechanics of the head-neck complex
is 10 cm the best pillow height
Does pillow height affect neck alignment?
What is the shoulder-width method for side-sleeper pillow height?
Can my pillow cause neck pain?
What pillow height range does research actually support?
Why do loft charts disagree with each other?
Key takeaways
- No pillow height study names a universal best height — the research supports ranges and individualization.
- X-ray study (n=16): 10 cm suggested for back sleeping; comfort unmeasured; the authors' own caveat applies.
- Computer model: high loft raised the cervical angle 66.4% and neck pressures significantly versus baseline.
- Side sleepers: shoulder-width-based heights of 9.7–11.8 cm gave the best cervical curve; comfort peaked slightly higher.
- Comfort and geometry can disagree by centimeters — weight both, and treat retail charts as starting points.
Last reviewed: 2026-11-01 by Honest Review Hub Editorial Team
References
- Kim, H. C., Jun, H. S., Kim, J. H., Ahn, J. H., Chang, I. B., Song, J. H., & Oh, J. K. (2015). The effect of different pillow heights on the parameters of cervicothoracic spine segments. Korean Journal of Spine, 12(3), 135–138. PMC4623167
- Ren, S., et al. (2016). Effect of pillow height on the biomechanics of the head-neck complex: A finite element study. PeerJ, 4, e2397. PMC5012320
- Beihang University. (n.d.). The individualized optimal pillow height and neck support design for side sleepers. Research portal record (year and journal not captured in our records — verification pending)
- Yamada, S., Hoshi, T., Toda, M., Tsuge, T., Matsudaira, K., & Oka, H. (2023). Changes in neck pain and somatic symptoms before and after the adjustment of the pillow height. Journal of Physical Therapy Science, 35(2), 106–113. PMC9889209
- Jiao, R., Xiao, W., Wang, M., Yu, S., & Li, H. (2025). The impact of pillow height on neck muscle activity: A pilot study. Sleep and Breathing, 29(1), 40. PubMed 39625641
- Airiti Library. (2021). Systematic review: Pillow effects on biomechanical properties, pain relief, and pulmonary function (1990–2019). Airiti Library record
- Pang, J. C. Y., Tsang, S. M. H., & Fu, A. C. L. (2021). Systematic review on the effectiveness of pillow interventions for neck pain. PubMed: 33895703
How we researched this page
We mapped 5 reader questions for this page from collected "people also ask" queries, then traced every number to 7 sources: a 2015 X-ray study (n=16), a 2016 finite-element model, an undated Beihang side-sleeper study (year/journal verification pending), a 2023 clinical adjustment study (n=84), a 2025 EMG pilot (n=15), a 2021 systematic review (1990–2019), and a 2021 35-trial review. Author names for the X-ray, adjustment and EMG studies were verified against the cited records in-session. No manufacturer claims were used as evidence; retail loft conventions are labeled as convention. Full method: how we research.