What the Studies Show About Pillow Height
Three small studies tested pillow height against neck alignment and pressure, and they agree on the direction but not on a number. Taller pillows raised measured neck pressure and extended the neck; a height band of 9.74 to 11.76 cm kept side sleepers' necks closest to their natural curve; and the review that surveyed the field since 1997 concluded no firm optimal height can yet be identified for back or side sleeping (Lei et al., 2021).
Applies to: anyone who wants the evidence behind pillow-height advice before acting on it.
Three studies. Ten people, sixteen people, nine people. That is the entire evidence base for one of the most-searched sleep questions on the internet — and it is the reason this page exists. The companion page gives you the practical test; this page gives you the studies themselves, one by one, with their designs, their numbers, and their limits, so you can see exactly how much weight the advice carries.
The headline is that the studies agree on direction and disagree on everything else. Height moves the neck: pressures rise, the neck extends, the modelled angle changes. But the studies measured different things in different people with different methods, the reviews built on them reach different conclusions, and the field's own survey says the optimal height cannot yet be identified. The most surprising finding is the comfort split in the side-sleeper study: the height that kept the neck healthiest was not the height people found most comfortable (Tian et al., 2025).
The short version: three studies, no settled number
If you want the one-paragraph version: taller pillows raise neck pressure and extend the neck; a band around 10 to 12 cm looks best for side sleepers' alignment; about 10 cm looks best for back sleepers' curve; and no review can turn those anchors into a prescription. The 2021 review of the field, surveying research published since 1997, concluded that the suggested range for optimal alignment, appropriate pressure distribution, and minimal muscle activity “cannot yet be identified considering the lack of sufficient evidence”, and that “there remain no firm conclusions about the optimal pillow height for the supine and lateral positions” (Lei et al., 2021).
The three anchor studies, at a glance. Ren et al. (2016): 10 healthy volunteers, pillows from 110 to 170 mm, pressures measured, alignment modelled — cervical pressure at 110 mm 65% lower than at 170 mm, modelled cervical angle up 66.4% (Ren et al., 2016). Kim et al. (2015): 16 asymptomatic adults aged 20 to 30, pillows of 0, 10, and 20 cm, X-ray measurements — the neck extended further back as height rose, and 10 cm was recommended for a normal cervical curve (Kim et al., 2015). Tian et al. (2025): 9 side sleepers, foam-latex pillows of 8, 10, 12, and 14 cm, height predicted from shoulder width — 9.74 to 11.76 cm kept the neck closest to its natural standing curve, 11.76 to 13.78 cm was rated most comfortable (Tian et al., 2025).
Two corrections to the record are built into this rewrite. First, an earlier version of this page's companion content quoted the Ren 65% figure backwards, as taller pillows causing “65% more pressure”; the paper reports the average cervical pressure at the lowest height as 65% lower than at the highest (Ren et al., 2016). Second, the Beihang side-sleeper study previously cited as undated is Tian et al. (2025), Medical & Biological Engineering & Computing 63(2):535–544. The 9.74–11.76 cm band itself was correct.
← Swipe to compare →| Study | Who | What it found | What it could not show |
|---|---|---|---|
| Ren 2016 | 10 healthy volunteers | Cervical pressure 65% lower at 110 mm than 170 mm; angle +66.4% modelled | Night-long sleep; people with neck pain |
| Kim 2015 | 16 asymptomatic adults, 20–30 | Neck extended with height; 10 cm recommended | Side sleeping; older or symptomatic people |
| Tian 2025 | 9 side sleepers | 9.74–11.76 cm best alignment; 11.76–13.78 cm most comfortable | Back sleeping; long-term outcomes |
Ren 2016: pressures measured, angles modelled
Ren et al. (2016), published in PeerJ, is the most-cited pillow-height study and the most misquoted. Ten healthy subjects, five men, average age 26, average height 167 cm and weight 59.6 kg, lay on pillows of four heights — 110, 130, 150, and 170 mm — while cranio-cervical pressure distribution was recorded and compared. Cervical spine alignment was studied with a finite element model built from Visible Human Project data, not measured on the volunteers (Ren et al., 2016). That distinction is the whole story of this study: the pressures are measurements; the alignment figures are a model.
The measured results: the average cranial pressure at 170 mm was approximately 30% higher than at 110 mm, and significantly different from the 130 and 150 mm heights; the average cervical pressure at 110 mm was 65% lower than at 170 mm, and significantly different from the 130 and 150 mm heights; the peak cervical pressures at 150 and 170 mm were significantly different from 110 mm (Ren et al., 2016). The modelled results: raising height from 110 to 170 mm caused an increase of 66.4% in cervical angle and 25.1% in lordosis distance, and a reduction of 43.4% in kyphosis distance (Ren et al., 2016). The authors' conclusion was that height elevation significantly increased cranial and cervical pressures and increased extension and lordosis of the cervical spine.
What the study cannot tell you: whether any of this matters over a night's sleep, in people with neck pain, on a real mattress, or at heights between the four tested. Ten young healthy volunteers lying still for minutes are not a population, and a finite element model of one anatomy is not a measurement. The study is also partly industry-linked — three of its authors were affiliated with a consumer products company — which is disclosed in the paper and worth knowing when weighing it. Its value is directional and quantitative: height moves pressure, steeply, and the effect is large enough to measure with ten people.
Kim 2015 and Tian 2025: two anchors, two positions
Kim et al. (2015), in the Korean Journal of Spine, is the back-sleeper anchor. Sixteen asymptomatic adults aged 20 to 30, with no history of injury, surgery, or treatment for neck symptoms, lay on pillows of 0, 10, and 20 cm while cervical sagittal parameters were measured with radiography. As height increased, the T1 slope and C2–7 Cobb's angle increased while neck tilt tended to decrease; the thoracic inlet angle stayed constant. The authors' conclusion, quoted exactly: “From the data obtained in this study, we recommend that the most suitable pillow height is 10 cm, considering the normal cervical lordosis” (Kim et al., 2015). Note the shape of that recommendation: picked from three options, in young adults without symptoms, for back sleeping, on the basis of a curve — not pain, not sleep quality, not a night's sleep.
Tian et al. (2025), in Medical & Biological Engineering & Computing, is the side-sleeper anchor — and the most sophisticated of the three. Nine healthy subjects lay laterally on foam-latex pillows of 8, 10, 12, and 14 cm, with and without neck support. Healthiness was evaluated from cervical spine morphology measured by motion capture and musculoskeletal internal force simulated by a multi-body model; comfort was evaluated by a standardised rating. The individualised height was identified by Hφ, calculated from the subject's shoulder width and the absolute pillow height, and the C1–T1 slope correlated with Hφ with R² of 0.80 for a flat pillow and 0.82 with neck support (Tian et al., 2025). The medium individualised height (Hφ 9.74–11.76 cm) with neck support showed a cervical curve closest to natural standing and the lowest musculoskeletal internal force; the sub-low height (Hφ 11.76–13.78 cm) with neck support showed the highest average comfort score (Tian et al., 2025).
The Tian study is the one that most rewards a close read, because it contains the field's central tension in miniature: the height that was best for the neck was not the height people liked most. Comfort correlated only moderately or weakly with the height measure, and the authors state plainly that comfort perception is not sensitive to it (Tian et al., 2025). A pillow that feels best on night one is not necessarily the pillow that holds the neck best — which is why the companion page tests height with a mirror and three mornings of scoring, not with a first impression.
What the reviews say — and where they disagree
The trials and reviews built on these studies do not speak with one voice. Pang et al. (2021), a systematic review and meta-analysis in Clinical Biomechanics, included 35 articles and pooled 9 high-quality studies with 555 participants: rubber pillows reduced neck pain (standardised mean difference −0.263, p < 0.001), waking pain (−0.228, p < 0.001), and neck disability (−0.506, p = 0.020), and raised satisfaction (1.144, p < 0.001), while pillow design did not influence sleep quality (0.047, p = 0.703). The authors' interpretation: side-lying alignment may depend on shape and height rather than material (Pang et al., 2021).
Ghosh et al. (2025), a systematic review in Rehabilitación, screened 29,091 records and included 5 studies with 239 participants with chronic neck pain: pain improvements were not statistically significant, and no pillow type was clearly superior. The two reviews disagree — one finds a small significant benefit, the other finds no significant benefit — because they asked different questions of different studies: Pang pooled high-quality trials of pillow design against waking symptoms; Ghosh restricted to chronic neck pain and a narrower set. Neither review tested height as an intervention. Both are reviews of pillow type, and height enters them only through the alignment discussion.
The trial evidence is thinner still. Gordon et al. (2010), a blinded field trial in 106 side sleepers who tried five pillows for a week each, found latex best and feather worst for waking symptoms, with a foam contour pillow no better than a foam regular pillow (Gordon et al., 2010). Gordon et al. (2019), a pilot randomised trial in 92 people with radiologically confirmed cervical spondylosis, found no pillow significantly altered any outcome measure, and concluded well-powered studies would need more than 400 symptomatic subjects (Gordon et al., 2019). The 2019 trial is the caution the whole field needs: the people most likely to buy a pillow for neck pain are the people in whom pillows have shown the least.
Lei et al. (2021) is the review that surveyed the height question directly, summarising research published since 1997 and organising the determinants as cervical alignment, body dimensions, contact pressure, and muscle activity. Its conclusion is the one this page is built around: the suggested range for optimal alignment, appropriate pressure distribution, and minimal muscle activity “cannot yet be identified considering the lack of sufficient evidence”, and “there remain no firm conclusions about the optimal pillow height for the supine and lateral positions” (Lei et al., 2021). Two of its authors were employees of a pillow technology company — disclosed in the paper. The conclusion stands regardless: the field has anchors, not answers.
What to do with three small studies
Read them as anchors, not prescriptions. The anchors are: about 10 cm for back sleeping (Kim, from three options in young adults); 9.74–11.76 cm for side sleepers' alignment, 11.76–13.78 cm for comfort (Tian, from nine people); and a steep pressure rise between 110 and 170 mm (Ren, from ten people). None of these is a number to buy. All of them point the same way: height matters, level matters, and the fit is personal — set by shoulder width and position, not by a chart.
The practical move is the companion page's: check your own pillow height with a mirror, adjust with a folded towel, and score three mornings. The by-position height numbers live in one place, the pillow loft guide, so the site does not maintain two competing sets of figures. And the firmness question — whether the pillow should be firm or soft — is a separate page with its own evidence: pillow firmness for neck pain.
Three rules for reading any new height claim you meet. First, ask who was studied: healthy volunteers or people with neck pain, and how many. Second, ask what was measured: pressure and alignment are not pain, and comfort is not alignment — Tian showed the two peaking at different heights (Tian et al., 2025). Third, ask whether the number is a measurement or a model: Ren's pressures were measured, its angles were modelled (Ren et al., 2016). A claim that passes all three tests is rare in this field, and worth more than ten that do not.
Try the rules on a claim you will meet in the wild: a pillow sold as “clinically proven at 11 cm”. Who was studied? The number resembles the Tian band, but Tian studied nine healthy side sleepers, not people with neck pain, and the band was individualised from shoulder width — 11 cm was inside the band for some people and outside it for others (Tian et al., 2025). What was measured? Alignment and comfort, not pain — and comfort peaked at a different height than alignment. Is it a measurement or a model? The alignment figures came from motion capture and simulation, not from a night's sleep. The claim fails all three tests: a band from nine people, presented as a fixed height, for an outcome it never measured. That is how most pillow-height marketing works, and the three rules catch it in under a minute.
Mistakes, myths, and what not to do
Myth: the studies prove a perfect height. They do not. Three small studies, three methods, and a review concluding no firm height can be identified (Lei et al., 2021). The anchors are real — 10 cm, 9.74–11.76 cm — but they are anchors from nine and sixteen people, not prescriptions.
Myth: taller pillows raise pressure 65%. The figure is real and the direction is backwards in the telling. Ren reported the average cervical pressure at 110 mm as 65% lower than at 170 mm (Ren et al., 2016). “65% more pressure” is the same comparison stated from the wrong end — and an earlier version of this site's content made exactly that error. Direction matters when the number is the argument.
Mistake: quoting the modelled angle as a measurement. The 66.4% rise in cervical angle from Ren is a finite element model, built from one person's anatomy — not a measurement on the ten volunteers (Ren et al., 2016). It is a legitimate model, and it is not the same kind of evidence as the pressure readings taken from those volunteers. Cite it as modelled, every time.
Mistake: treating comfort as evidence. In the side-sleeper study, the most comfortable height was not the healthiest height (Tian et al., 2025). A pillow that feels right on night one can still hold the neck wrong all night. Comfort is a vote, not a verdict.
Myth: the reviews agree. Pang found a small significant benefit from pillow design; Ghosh found no significant benefit (Pang et al., 2021) (Ghosh et al., 2025). They asked different questions of different studies. Quoting one without the other is cherry-picking, and this page quotes both.
What the evidence cannot tell you
Almost everything a reader actually wants to know. No study in this review measured pain as an outcome of height: Ren measured pressure and modelled alignment, Kim measured X-ray angles, Tian measured morphology, simulated force, and comfort (Ren et al., 2016) (Kim et al., 2015) (Tian et al., 2025). No study followed people over nights, let alone weeks. No study tested height in people with neck pain — the volunteers were healthy or asymptomatic — and no study tested height against mattress firmness, which changes the gap the pillow must fill. The reviews disagree with each other, and the field's own survey says the optimal range cannot yet be identified (Lei et al., 2021) (Pang et al., 2021) (Ghosh et al., 2025).
Two of the anchor studies carry industry links that should be stated: Ren's team included authors affiliated with a consumer products company, and two of Lei's authors were employees of a pillow technology company. Both disclosures are in the papers. They do not invalidate the findings — the pressure measurements are measurements — but they are part of the weight you put on them.
What would settle the height question is a trial that randomises people with morning neck pain to heights chosen by the mirror check versus their usual pillow, measures morning pain and sleep quality over four weeks, and records position with accelerometers. No such trial exists. The honest statement of the field, in the field's own words, is Lei's: the range “cannot yet be identified” (Lei et al., 2021). This page will be updated when that changes.
When the studies do not apply
These studies do not apply to necks with a diagnosis. The volunteers were healthy or asymptomatic; nobody studied height in cervical spondylosis, arthritis, or radiculopathy — and the one pilot trial in people with confirmed cervical spondylosis found no pillow significantly altered any outcome (Gordon et al., 2019). If your neck pain comes with arm numbness or weakness, fever, pain after a serious injury, or symptoms that keep worsening, the studies on this page are not your guide: see a clinician promptly, and see how to sleep with cervical radiculopathy for the nerve-pain picture.
The studies also do not apply to the question most readers actually ask, which is “what should I do tonight?” That is the companion page's job: check your own pillow height. This page answers what the studies show. The two pages are kept separate deliberately, so neither has to pretend to be the other.
Key takeaways
- Taller pillows raised measured neck pressure steeply — cervical pressure at 110 mm was 65% lower than at 170 mm in 10 volunteers (Ren et al., 2016).
- About 10 cm was the back-sleeper anchor from an X-ray study of 16 adults (Kim et al., 2015); 9.74–11.76 cm was the side-sleeper alignment band from 9 people (Tian et al., 2025).
- The healthiest height was not the most comfortable height in the side-sleeper study (Tian et al., 2025).
- The reviews disagree: Pang found a small significant benefit from pillow design; Ghosh found no significant benefit (Pang et al., 2021) (Ghosh et al., 2025).
- No firm optimal height can yet be identified for back or side sleeping — the field's own review says so (Lei et al., 2021).
- The 66.4% angle rise is modelled, not measured — cite it as modelled, every time (Ren et al., 2016).
Questions people actually ask
What do studies say about pillow height?
Is 10 cm the best pillow height?
What pillow height did the Beihang study find best?
Do taller pillows raise neck pressure?
Why do studies disagree about pillow height?
Is there a perfect pillow height for everyone?
Keep exploring: your next question
References
- Pang, J. C. Y., Tsang, S. M. H., & Fu, A. C. L. (2021). The effects of pillow designs on neck pain, waking symptoms, neck disability, sleep quality and spinal alignment in adults: a systematic review and meta-analysis. Clinical Biomechanics, 85, 105353. PubMed: 33895703
- Ghosh, S., Goyal, M., & Goyal, K. (2025). Effect of pillow on pain, disability and sleep quality in patients with chronic neck pain: a systematic review. Rehabilitación, 59(3), 100922. PubMed: 40633255
- Gordon, S. J., Grimmer-Somers, K. A., & Trott, P. H. (2010). Pillow use: the behavior of cervical stiffness, headache and scapular/arm pain. Journal of Pain Research, 3, 137–145. PubMed: 21197317
- Gordon, S. J., Grimmer, K. A., & Buttner, P. (2019). Pillow preferences of people with neck pain and known spinal degeneration: a pilot randomized controlled trial. European Journal of Physical and Rehabilitation Medicine, 55(6), 783–791. PubMed: 31489809
- Ren, S., et al. (2016). Effect of pillow height on the biomechanics of the head-neck complex. PeerJ, 4, e2397. PubMed: 27635354 / PMC5012320
- Tian, S., et al. (2025). The individualized optimal pillow height and neck support design for side sleepers. Medical & Biological Engineering & Computing, 63(2), 535–544. PubMed: 39412632
- Kim, H. C., et al. (2015). The effect of different pillow heights on the parameters of cervicothoracic spine segments. Korean Journal of Spine, 12(3), 135–138. PubMed: 26512267
- Lei, J. X., et al. (2021). Ergonomic consideration in pillow height determinants and evaluation. Healthcare, 9(10), 1333. PubMed: 34683013
Last updated: 2026-11-21 by Adeel Farooq
Published 21 November 2026 — new page (gap G17). Images reuse existing neck-pain assets; dedicated photography to be added later.
How we researched this page
We mapped 6 reader questions for this page from collected “people also ask” queries and live search results for “what do studies say about pillow height” (21 November 2026), then traced every claim to 8 sources. This page is the retargeted evidence half of the pillow-height pair: it carries no primary keyword of its own, and the practical test lives on the companion self-check page while the by-position numbers live in the loft guide. Two corrections are built in: the Ren 65% figure is stated the right way round, and the Beihang study is identified as Tian et al. (2025). No manufacturer claims were used as evidence. Counts verified November 2026. Full method: how we research.