Sleep · Metric

REM Sleep

Rapid eye movement sleep — the stage most associated with dreaming, emotional processing, and procedural memory consolidation. Target: 20–25% of total sleep time.

This page will help you understand what your device's REM percentage is actually estimating and how it should shift with age, so you can judge whether your own number is unusual or expected.

Save your numbers

Enter your own REM sleep percentage below to save it in this browser and compare it with the reference value. No account, no sign-up — the value stays on your device only.

What affects this the most?

Alcohol
Strong suppressant, early night
Antidepressants
SSRIs significantly suppress REM
Sleep deprivation
Early waking cuts REM
Inconsistent schedule
Disrupts REM timing

Definition

REM (rapid eye movement) sleep is characterized by rapid lateral eye movements, muscle atonia, and brain activity resembling wakefulness. Most vivid dreaming occurs during REM. The brain is highly active — glucose metabolism during REM approaches waking levels.

REM sleep concentrates in the second half of the night. Each sleep cycle (approximately 90 minutes) ends with a REM period, and REM periods lengthen as the night progresses. Cutting sleep short disproportionately reduces REM.

Wearables can't see brain state directly: Consumer devices infer REM from heart rate, HRV, and movement patterns rather than the eye-movement and brainwave signals a sleep lab uses. They're reasonably good at detecting that REM occurred, less precise at pinpointing exact duration.

Functions of REM sleep

Normal REM percentage by age

Age GroupREM % of Total SleepMinutes (8h sleep)
18–2520–25%96–120 min
26–4019–24%91–115 min
41–5518–22%86–106 min
56–7016–20%77–96 min
71+14–18%67–86 min

Source: Ohayon MM et al. Sleep. 2004.

What reduces REM

Oura Ring
REM detection — validated in multiple studies
Garmin
REM in sleep report. Moderate accuracy.
Whoop
REM reported in sleep breakdown
Apple Watch
REM in iOS Sleep app (iOS 16+)
Accuracy note: Wearable REM detection shows moderate PSG agreement (sensitivity ~70%). Better at detecting REM presence than precise duration.

Myths vs. reality

MythReality
"More REM is always better"REM has a healthy range; unusually high reported REM is often a detection artifact rather than a sign of superior sleep.
"A low-REM night means I didn't dream"Dreaming can occur in other sleep stages too, and a single night's REM percentage doesn't reliably indicate dream recall.
"My REM % should be identical every night"REM naturally varies night to night with sleep duration, timing, alcohol, and stress — trend matters more than any single reading.
"REM percentage is comparable across brands"Different detection algorithms mean the same sleep can register different REM totals on different devices.

How each device measures this

Oura Ring
Detects REM from characteristic heart rate and breathing irregularity patterns plus stillness; not a substitute for EEG-based staging.
Whoop
Identifies REM windows using heart rate variability shifts; brand-to-brand disagreement on exact REM timing is common.
Garmin
Wrist PPG and motion sensing infer REM periods, generally with wider error margins than ring-based optical sensors.
Apple Watch
Reports REM as one of several sleep stages using motion and heart rate; treat night-to-night REM totals as trend data, not exact measurement.
Not medical advice: Data presented here is for educational reference only. Consult a qualified clinician for health concerns.