Rapid eye movement (REM) sleep is one of the two broad sleep states, alongside non-REM sleep. It is a normal part of sleep, not a separate kind of rest that you need to optimize on its own. In a sleep laboratory, REM is identified from a combination of brain activity, eye movements, and muscle tone. Dreaming or a consumer-device score cannot identify it with the same certainty 1.
REM sleep returns several times during a sleep period and is usually more prominent later in that period. That pattern matters more than hitting a fixed number of REM minutes or following a rigid 90-minute clock 2.
What happens during REM sleep?
REM sleep combines features that do not usually occur together while awake or in non-REM sleep:
- Brain activity becomes relatively activated. The scalp electroencephalogram, or EEG, shows low-amplitude, mixed-frequency activity that can look closer to wakefulness than deep non-REM sleep.
- The eyes may move rapidly. These movements occur in bursts, so the eyes do not need to move throughout the whole REM period.
- Most large skeletal muscles have very low tone. This REM atonia limits many limb and trunk movements. It does not stop breathing, eye movement, or every small muscle twitch.
- Breathing and heart rate can become more variable. A change in either signal alone is not enough to identify REM because many other factors can alter them.
Researchers describe REM as a collection of brain, eye, muscle, and autonomic features rather than a state defined by one sign 1. This is why a vivid dream, a twitch, or a change in heart rate cannot confirm that someone was in REM.
When does REM occur?
Sleep usually begins in non-REM sleep and moves between non-REM and REM several times. The National Heart, Lung, and Blood Institute describes sleep cycles as commonly restarting in a broad range of roughly 80 to 100 minutes, with more REM later in the sleep period 2. That average is not a timer. Cycle length and the time to the first REM period vary between people and from one night to another.
The later distribution also explains why an early alarm or a shortened sleep opportunity may remove part of a REM-rich portion of sleep. It does not mean everyone must sleep on the same schedule. The relevant pattern follows the person's sleep period, whether that occurs at night or at another biologically appropriate time.
Sleep architecture also changes across the lifespan. Newborns spend more time in REM than adults, and pooled sleep-study data show that REM measures continue to vary with age 23. These group patterns are not personal targets.
Is REM the same as dreaming?
No. Dream reports are more common after awakenings from REM, and REM dreams are often described as vivid or story-like. People also report dreams after awakenings from non-REM sleep. Researchers still have to rely largely on a person's report after waking, which makes the exact timing and function of a dream difficult to establish 4.
For practical interpretation:
- Remembering a dream does not prove how long you spent in REM.
- Not remembering a dream does not show that you missed REM.
- Vivid or unusual dreams do not, by themselves, diagnose a sleep disorder.
The purpose of dreaming remains an open scientific question. It should not be presented as a proven method of emotional repair, creative problem-solving, or memory storage 4.
What does REM do for memory and emotion?
Sleep supports memory, but the evidence does not assign memory to REM alone. A major review of sleep and memory found that both non-REM and REM processes are involved across different forms of memory, and that the sequence between stages may matter 5. It is therefore misleading to describe REM as the stage that simply transfers every new memory into long-term storage.
REM is also studied in emotional learning and affective processing. The evidence is less settled than many summaries suggest. A 2024 systematic review found substantial differences in REM-disruption methods and only a small number of human experiments, most involving acute disruption over one night 6. Forced awakenings can also introduce stress and sleep fragmentation, which makes it harder to isolate the effect of REM itself.
The evidence supports a narrower conclusion: REM probably participates in memory and emotional processing as part of a full sleep period. Current evidence does not show that selectively increasing a person's REM percentage will improve memory, mood, or creativity.
How is REM sleep measured?
A laboratory polysomnogram uses several signals at the same time. The core signals for sleep staging are:
- EEG for electrical activity at the scalp
- electrooculography, or EOG, for eye movements
- electromyography, or EMG, for muscle activity, including chin muscle tone
A clinical study may also record airflow, breathing effort, oxygen level, heart rhythm, limb movement, and video. Those additional signals help explain why REM is interrupted or why movement occurs. A trained scorer reviews the signals together rather than diagnosing REM from eye movement alone 1.
A sleep-study report may include REM latency, time in REM, the share of sleep scored as REM, awakenings, respiratory events, and muscle activity. Clinicians interpret these measures alongside age, sleep timing, total sleep time, medicines, substances, symptoms, and the rest of the study. A single REM percentage is not a diagnosis.
Can a watch or ring measure REM accurately?
Consumer devices can estimate sleep stages, but most do not record the same EEG, EOG, and chin EMG combination used for clinical sleep staging. Their algorithms infer stages from signals such as movement, pulse, and pulse variability 7.
In a laboratory comparison of seven consumer trackers, performance differed by device and sleep measure. The study involved young adults without diagnosed sleep or major health disorders, so its results do not establish accuracy for every user or for people with fragmented sleep 8.
The American Academy of Sleep Medicine states that consumer sleep technology should not be used to diagnose or treat a sleep disorder unless it has been appropriately cleared and tested for that purpose 7. A tracker can be useful for observing broad routines or bringing a recurring concern to a clinician. It should not be used to decide that you have REM deprivation, REM sleep behavior disorder, narcolepsy, or sleep apnea.
What if my tracker says I get too little REM?
Do not try to correct one low reading. First consider whether the device captured the whole sleep period and whether the result is part of a consistent pattern. Even repeated estimates need to be interpreted with symptoms, total sleep opportunity, age, medicines, alcohol use, and possible sleep disruption 37.
If you feel well and have no concerning symptoms, the number alone is not evidence of a disorder. If you are persistently sleepy, have breathing symptoms, or physically act out dreams, seek evaluation based on those symptoms rather than the score.
What can change REM sleep?
Sleep timing and REM rebound
Because REM is generally more prominent later in a sleep period, changing the length or timing of sleep can change how much REM is recorded. After REM has been restricted or suppressed, later sleep can sometimes show more REM or a stronger tendency to enter REM. This response is called REM rebound 9.
REM rebound is a description of sleep architecture, not a wellness goal. Vivid dreams after sleep loss, a medicine change, or stopping a substance do not show that the brain has completed a measurable amount of repair.
Medicines and alcohol
Many medicines affect sleep architecture, but their effects are not interchangeable. Reviews of antidepressants show that effects depend on the specific medicine, dose, timing, and duration of treatment. Many SSRIs and SNRIs delay or reduce REM, while other antidepressants have different patterns 10.
A lower REM measure while taking a prescribed medicine does not show that the medicine is damaging the brain or failing to work. Do not skip, taper, or stop a medicine to change a tracker score. Discuss new vivid dreams, dream enactment, severe sleep disruption, or concerning daytime symptoms with the prescriber.
Alcohol is not a reliable way to improve sleep. A systematic review and meta-analysis of 27 controlled studies in healthy adults found that presleep alcohol delayed the first REM period and reduced REM duration, with greater disruption at higher doses 11. The studies did not establish one safe bedtime cutoff that applies to everyone. See alcohol and sleep for the broader effects.
REM sleep behavior disorder
In REM sleep behavior disorder, normal REM atonia is reduced and a person may repeatedly talk, punch, kick, grab, or leave the bed while apparently enacting a dream. Dream enactment can have other causes, including sleep-disordered breathing, medicines, and other parasomnias. A formal diagnosis requires a detailed history and video polysomnography showing REM sleep without normal atonia while other explanations are assessed 12.
Recurrent dream enactment deserves medical evaluation even if it is infrequent. Until it is assessed, remove weapons and sharp or breakable objects from the bedroom, move or pad hard furniture near the bed, place a soft mat beside the bed, and sleep separately if a bed partner could be injured 13.
Narcolepsy and sleep paralysis
Narcolepsy disrupts control of sleep and wakefulness. Its central symptom is excessive daytime sleepiness. Type 1 narcolepsy also includes cataplexy, a brief loss of muscle control triggered by emotion. Sleep paralysis and vivid experiences while falling asleep or waking can occur in narcolepsy, but either symptom can occur without narcolepsy 14.
Clinicians may investigate narcolepsy with an overnight polysomnogram followed by a multiple sleep latency test, which measures how quickly someone falls asleep during scheduled naps and whether REM begins unusually soon. Sleep schedule, insufficient sleep, other disorders, medicines, and substances can affect the result, so testing must follow a standardized protocol and be interpreted clinically 15.
A brief episode of sleep paralysis is not proof that your REM sleep is abnormal. Seek care if episodes are frequent or distressing, especially when they occur with uncontrollable daytime sleepiness or cataplexy.
Sleep apnea
Obstructive sleep apnea can occur in both non-REM and REM. In some people, breathing events cluster in REM or become longer and cause greater oxygen drops during REM 16. Those events can fragment sleep, but a consumer report of low REM cannot diagnose their cause.
Breathing that repeatedly stops and starts, loud habitual snoring, gasping, and persistent daytime sleepiness are reasons to discuss sleep apnea with a healthcare professional 17. Diagnosis depends on an appropriate sleep test, not a stage estimate from a watch.
Can you selectively increase REM sleep?
Current evidence does not establish a method that lets a healthy person selectively increase REM to a personal target 9. Trying to manipulate one stage can distract from the factors that matter more:
- Allow enough time for a full sleep period.
- Keep sleep timing reasonably consistent when your schedule allows.
- Do not use alcohol as a sleep aid.
- Do not change prescribed medicine based on a tracker.
- Treat symptoms and diagnosed disorders rather than chasing a stage score.
These steps support sleep as a whole. They do not promise a particular REM percentage, and more REM is not automatically better.
When to seek medical care
Arrange an evaluation if you or a bed partner notice:
- repeated punching, kicking, grabbing, shouting, or leaving the bed during dreams
- injuries or near injuries during sleep
- breathing pauses, choking, gasping, or loud habitual snoring
- persistent daytime sleepiness despite adequate sleep opportunity
- sudden sleep episodes or emotion-triggered muscle weakness
- a major sleep change after starting, changing, or stopping a medicine or substance
Do not drive when you are struggling to stay awake. Drowsiness reduces attention and reaction time, and the National Highway Traffic Safety Administration advises delaying driving until you are well rested 18. Dangerous or uncontrollable sleepiness needs prompt medical assessment.
FAQs
How much REM sleep do adults need?
There is no prescribed REM-minute or REM-percentage target for an individual. Sleep architecture varies with age and context, and clinical interpretation depends on the whole sleep study and the person's symptoms 3.
Is more REM always better?
No. More REM can reflect normal later sleep, recovery after suppression, medication or substance changes, or a clinical pattern that needs context. A higher number is not automatically a sign of better restoration 9.
Do vivid dreams mean I had a lot of REM?
Not reliably. Vivid dreams are often reported from REM, but dreams can occur in non-REM sleep and recall depends partly on when and how you wake 4.
Can a sleep tracker diagnose REM sleep behavior disorder?
No. REM sleep behavior disorder requires clinical assessment and video polysomnography that evaluates behavior and muscle tone during confirmed REM sleep 12.
Should I change a medicine if it lowers my REM score?
No. Medicines can change REM without making a consumer score clinically meaningful. Discuss symptoms and concerns with the prescriber, and never stop or taper a medicine solely to change sleep-stage data 10.
