Deep Sleep and REM Scores
What the deep sleep and REM figures on sleep trackers mean, why there is no ideal percentage for everyone, and what research does and does not show about them.
Content last updated . This page has not been medically reviewed.
What these numbers are
Many sleep apps show minutes or percentages of "deep" and "REM" sleep, and some fold them into a single sleep score. Deep sleep corresponds to stage 3 of non-REM sleep, which occurs in longer periods in the first half of the night. REM (rapid eye movement) sleep first occurs about 90 minutes after falling asleep, and most dreaming happens in it. [1]In a lab, these stages are defined by brain waves recorded during polysomnography. A wrist or ring device has no brain-wave sensor, so its stage figures are estimates from movement and heart signals, and any "score" built from them is a company's own calculation. Labels used here: established, promising but limited, mixed, insufficient evidence, marketing claim and Proper evidence not found (we looked and found nothing usable).
No ideal percentage for everyone
No single deep sleep or REM percentage is right for every person, and we found no official target you can apply to a device reading. Sleep stages change across life. In a meta-analysis of 65 studies of healthy people aged 5 to 102, the percentage of slow-wave (deep) sleep and of REM sleep decreased with age in adults, and the size of these effects shifted depending on how carefully participants had been screened for illness and sleep disorders. [2] NINDS also notes that people spend less of their time in REM sleep as they age. [1] A "deep sleep" number that looks low for a 20-year-old may be unremarkable for a 70-year-old, and a device does not compare you with the right group.Individual nights vary too. A good week and a bad week on the same device are not proof of a change in your health. See sleep trends vs individual nights.
How accurate are the figures?
Accuracy is device-specific. In a 35-person overnight lab study, the Oura Ring Gen3 was not significantly different from polysomnography on deep or REM minutes, the Fitbit Sense 2 underestimated deep sleep by about 15 minutes, and the Apple Watch Series 8 underestimated deep sleep by about 43 minutes. [3] In a 2026 home study, the Apple Watch Series 7 overestimated REM by about 30 minutes, and agreement across three devices was generally weak with wide limits. [4] A six-device study found kappa agreement with polysomnography of 0.21 to 0.53 across the devices. [5] Full tables are on sleep-stage estimates: how accurate are they?. Label for stage figures: mixed.What research says about deep sleep and REM and health
Lab-measured REM has been studied. In two cohorts that used polysomnography, a lower percentage of REM sleep was associated with a greater risk of death over follow-up, and in the older-men cohort those with under 15% REM had a higher mortality rate than those with 15% or more. [6] This is observational research in specific groups (the main one was older men), done with full sleep studies, not wearables. It cannot show that raising REM would lower risk. Label: insufficient evidence for any action you could take.We found no trials showing that raising a wearable's deep or REM figure improves health or performance: Proper evidence not found (we searched PubMed for intervention trials on deep sleep and wearable-derived stage data). Claims that a product "increases deep sleep" are a marketing claim unless supported by lab-measured trials in people, and a change in a biomarker is not the same as a change in how you feel or live. For a related example of product claims, see sound-based sleep claims.
What can safely be said
The American Academy of Sleep Medicine says consumer sleep technology cannot be used to diagnose or treat sleep disorders. [7] Clinicians have described people who sought treatment because of "light" or restless sleep on tracker data, and who pursued an ideal night. [8] A stage score is best treated as a rough trend marker at most.Who should be careful
If low scores make you anxious or keep you awake, see orthosomnia and sleep-tracking anxiety. If you have long-lasting poor sleep, daytime sleepiness, loud snoring or pauses in breathing, ask a clinician; a device cannot test for these. For habits that are better supported, see sleep hygiene.Frequently asked questions
How much deep sleep should I get?
There is no figure that fits everyone, and deep sleep decreases with age in healthy adults. [2]
Is my REM percentage too low?
A device cannot say. Its REM figure is an estimate, and lab-measured low REM is linked with higher mortality only in observational studies. [6]
Can I increase my deep sleep score?
We found no trials showing that raising the score improves health. Sleep habits are the better-supported starting point. [9]
Is the sleep score a medical measurement?
No. It is a proprietary calculation, and consumer devices are not for diagnosing sleep disorders. [7]
References
- National Institute of Neurological Disorders and Stroke (NINDS). Brain Basics: Understanding Sleep. Read 2026-10-22.
- Meta-analysis of quantitative sleep parameters from childhood to old age in healthy individuals. Sleep 2004. Read 2026-10-22.
- Accuracy of Three Commercial Wearable Devices for Sleep Tracking in Healthy Adults. Sensors 2024. Read 2026-10-22.
- Performance of three consumer sleep-tracking devices compared with Actigraphy and Polysomnography. Sleep Medicine 2026. Read 2026-10-22.
- A performance validation of six commercial wrist-worn wearable sleep-tracking devices for sleep stage scoring compared to polysomnography. Sleep Advances 2025. Read 2026-10-22.
- Association of Rapid Eye Movement Sleep With Mortality in Middle-aged and Older Adults. JAMA Neurol 2020. Read 2026-10-22.
- Consumer Sleep Technology: An American Academy of Sleep Medicine Position Statement. J Clin Sleep Med 2018. Read 2026-10-22.
- Orthosomnia: Are Some Patients Taking the Quantified Self Too Far? J Clin Sleep Med 2017. Read 2026-10-22.
- NHS. Insomnia. Page last reviewed 19 March 2024. Read 2026-10-22.