You Sleep 7 Hours.
But Are You Getting Enough REM?
Most people track total sleep hours. Almost nobody tracks when their alarm cuts in. The problem: REM sleep concentrates heavily in your last 1–2 cycles. A 6:00am alarm instead of 7:30am does not cost you 20% of your REM — it costs you roughly 30%. Walker (2017, Why We Sleep) identified REM as central to emotional memory processing, creative problem-solving, and mood regulation. If you wake tired even after enough total hours, the missing variable is likely cycle completion — not sleep duration. This calculator shows you exactly what you have, and what you are losing.
Adults need roughly 90–120 minutes of REM sleep per night — approximately 20–25% of 7.5–8 hours of total sleep (AASM, 2007; Hirshkowitz et al., 2015). REM is not evenly spread: the first cycle produces only ~10 minutes of REM, while the fifth cycle produces ~45 minutes. This means losing the last 90-minute cycle costs about 30% of total nightly REM while removing only ~17% of total sleep time. Consistently low REM — from short sleep, alcohol, or certain medications — is linked to impaired emotional regulation, reduced learning consolidation, and next-day mood disruption.
Estimate Your Nightly REM Sleep
Move the slider to your typical sleep duration and select your age group. Estimates update live and use AASM population averages — individual REM varies.
REM Per Sleep Cycle — Why Duration Matters
REM is not evenly distributed. The brain prioritises slow-wave deep sleep in early cycles and progressively shifts to REM-dominant cycles as the night continues. The fifth cycle alone contains as much REM as the first three combined (Carskadon & Dement, 2011). Adjust the slider above — cycles beyond your sleep duration are shown as missed.
6 hours (4 cycles) ≈ 100 min REM · 7.5 hours (5 cycles) ≈ 145 min REM
Just 1.5 extra hours = ~45% more REM sleep.
What Reduces Your REM Sleep
These are the most evidence-backed REM suppressors. Each one acts via a different biological mechanism — meaning multiple factors compound independently rather than averaging out.
-
Alcohol (even 1–2 units within 3h of bed) Strong Evidence
Suppresses REM in the first half of the night by increasing slow-wave sleep, then causes a REM rebound in the second half — producing fragmented, lighter sleep and often vivid or disturbing dreams. The net result is structurally disrupted REM regardless of how early drinking stopped.
Ebrahim et al. (2013), Alcoholism: Clinical and Experimental Research — systematic review of 27 studies.
-
Most SSRIs / SNRIs Strong Evidence
Antidepressants in these classes significantly suppress REM — a documented side effect across the drug class. Never stop medication to improve REM. If this concerns you, discuss options with your prescribing doctor. Some agents have less REM-suppressive profiles than others.
Wilson & Argyropoulos (2005), CNS Drugs — antidepressants and sleep architecture review.
-
Cannabis (THC) Strong Evidence
THC suppresses REM acutely via CB1 receptor activity. Regular users often report minimal dreaming. REM rebound is common and can be intense on cessation — expect vivid, sometimes disturbing dreams for 2–4 weeks after stopping regular use.
Bhatt et al. (2020), Current Psychiatry Reports — cannabis and sleep architecture.
-
Chronic sleep restriction Strong Evidence
The single largest cause of inadequate REM is simply not sleeping long enough. Because REM concentrates in cycles 4 and 5, cutting sleep by 90 minutes removes ~30% of REM while removing only ~17% of total sleep time — a disproportionate loss. This is the mechanism behind why 6-hour sleepers can feel cognitively adequate yet test significantly worse on emotional and creative tasks.
Van Dongen et al. (2003), Sleep; Walker (2017), Why We Sleep.
-
Very early sleep onset vs. natural phase Moderate Evidence
REM timing is partly controlled by circadian clock signals rather than purely homeostatic sleep pressure. Sleeping significantly earlier than your natural circadian phase shifts the REM-heavy portion of sleep outside the window it normally occurs, reducing overall REM quality and density even when total hours appear adequate.
Czeisler et al. (1980), Science — circadian timing and REM distribution.
-
Hot sleep environment (>20°C / 68°F) Moderate Evidence
Core body temperature naturally drops during REM. A sleep environment warmer than the optimal 16–19°C (60–67°F) range impairs this thermoregulatory drop, reducing REM quality. Walker (2017) notes this as one of the most underappreciated environmental REM suppressors for urban sleepers.
Okamoto-Mizuno & Mizuno (2012), Journal of Physiological Anthropology.
Frequently Asked Questions
How much REM sleep do I need per night?
Adults typically need 90–120 minutes of REM per night, representing 20–25% of 7.5–8 hours total sleep (AASM 2007; Hirshkowitz et al., 2015). REM occurs in 4–6 discrete periods that grow progressively longer in later cycles. The first cycle produces roughly 10 minutes of REM. The fifth cycle produces roughly 45 minutes.
Consistently falling below 90 minutes — whether from short sleep, alcohol, or certain medications — is associated with impaired emotional regulation, reduced learning consolidation, and weakened creative thinking (Walker, 2017). The threshold is not a cliff edge: small, consistent deficits accumulate over weeks rather than causing acute impairment overnight.
Why do I dream more in the morning?
REM sleep concentrates in later cycles and is partially controlled by your circadian clock rather than purely by how long you have been asleep. REM propensity peaks in the early morning hours — roughly between 5 AM and 8 AM for most adults — which is why the most vivid and memorable dreams almost always occur just before waking (Carskadon & Dement, 2011).
An alarm that cuts off the last 60–90 minutes of a natural sleep window removes precisely this REM-dense period. This is why many people who use alarms report rarely remembering dreams, while the same people on weekend mornings without alarms may recall vivid dream sequences.
Can I make up lost REM sleep?
Partially. After REM deprivation, the brain increases REM pressure — shown as REM rebound on the following nights. However, the rebound does not fully restore all missed REM. The practical implication from Van Dongen et al. (2003) is that preventing REM loss is significantly more effective than recovering it.
The most reliable recovery strategy is 2–3 nights of adequate, undisturbed sleep without an early alarm — allowing the natural REM-dense morning window to complete fully. This is why sleep researchers consistently recommend protecting the last 90 minutes of sleep rather than adding an extra hour to the beginning of the night.
What is the difference between REM and deep sleep?
Deep sleep (N3, slow-wave sleep) dominates the first half of the night and is primarily associated with physical recovery, immune function, and declarative memory consolidation. REM sleep dominates the second half and is associated with emotional memory processing, procedural memory, creative integration, and mood regulation.
Both are essential — they serve different biological functions and cannot substitute for each other. People who sleep 6 hours typically lose more REM than deep sleep proportionally, while people who sleep 10 hours gain more REM than deep sleep relative to their baseline. Walker (2017) describes this as the brain using the night like a two-stage filing system: deep sleep archives facts, REM contextualises and connects them.
Is my Fitbit or Apple Watch REM reading accurate?
Consumer wearables achieve approximately 70–78% accuracy for REM detection compared to clinical polysomnography (de Zambotti et al., 2019, Sleep Medicine Reviews). That is meaningful — but it also means roughly 1 in 4 REM classifications is incorrect. Wearables tend to overestimate light sleep and underestimate both deep sleep and REM.
Use wearable REM data as a directional trend signal over weeks, not as a precise nightly measurement. If your wearable consistently shows very low REM (under 60 minutes nightly) despite adequate total sleep, it is worth investigating further — but confirm with a GP rather than a consumer device alone.
Does REM sleep decrease with age?
Yes. REM percentage decreases modestly with age. Infants spend approximately 50% of sleep in REM; adults 18–64 average 20–25%; adults 65+ typically average 15–20% (Ohayon et al., 2004, Sleep). The reduction reflects broader changes in sleep architecture with aging, including reduced slow-wave sleep and more frequent night awakenings.
Older adults also tend to experience sleep timing shifts — earlier bedtimes and wake times — which can misalign their REM window relative to when they are actually in bed. The age group selector in the calculator above accounts for these population-level differences.
What is REM Sleep Behaviour Disorder (RBD)?
RBD is a clinical condition in which the normal muscle paralysis (atonia) that occurs during REM sleep fails — causing people to physically act out their dreams, sometimes injuring themselves or a bed partner. It is diagnosed via polysomnography, not consumer wearables. RBD is associated with neurodegenerative conditions including Parkinson’s disease and Lewy body dementia, and can precede motor symptoms by years.
If you or a partner physically move, speak, or seem to be acting out dreams during sleep, consult a sleep specialist or neurologist. RBD is not the same as normal sleep talking or occasional jerking movements — it involves complex, often violent dream-enacting behaviour. This calculator does not screen for RBD.
Sources
- American Academy of Sleep Medicine (2007). The AASM Manual for the Scoring of Sleep and Associated Events. Rules, Terminology and Technical Specifications.
- Hirshkowitz M et al. (2015). “National Sleep Foundation’s sleep time duration recommendations.” Sleep Health, 1(1):40–43.
- Carskadon MA & Dement WC (2011). “Normal human sleep: an overview.” In Kryger MH et al. (eds), Principles and Practice of Sleep Medicine, 5th ed.
- de Zambotti M et al. (2019). “Wearable sleep technology in clinical and research settings.” Sleep Medicine Reviews, 48:101453.
- Walker MP (2017). Why We Sleep: Unlocking the Power of Sleep and Dreams. Scribner.
- Ebrahim IO et al. (2013). “Alcohol and sleep I: effects on normal sleep.” Alcoholism: Clinical and Experimental Research, 37(4):539–549.
- Van Dongen HPA et al. (2003). “The cumulative cost of additional wakefulness.” Sleep, 26(2):117–126.
- Wilson S & Argyropoulos S (2005). “Antidepressants and sleep: a qualitative review of the literature.” CNS Drugs, 19(4):307–324.
- Ohayon MM et al. (2004). “Meta-analysis of quantitative sleep parameters from childhood to old age.” Sleep, 27(7):1255–1273.
- Bhatt DL et al. (2020). “Cannabis and sleep.” Current Psychiatry Reports, 22(4):18.
- Okamoto-Mizuno K & Mizuno K (2012). “Effects of thermal environment on sleep and circadian rhythm.” Journal of Physiological Anthropology, 31(1):14.