Most adults enter their first period of REM sleep about 60 to 120 minutes after falling asleep. You don’t go straight into REM. Instead, your brain moves through three progressively deeper stages of non-REM sleep first, and only then shifts into the lighter, dream-heavy REM state. That first full cycle from sleep onset through REM typically takes 90 to 120 minutes.
What Happens in the 90 Minutes Before REM
Sleep unfolds in stages, and each one has to run its course before REM begins. Stage 1 is the lightest phase, lasting only a few minutes as you transition from wakefulness. It accounts for roughly 5% of your total sleep time. Stage 2 is a longer stretch of light sleep where your heart rate slows and body temperature drops. Stage 3 is deep sleep, the most physically restorative phase, and it dominates the first cycle of the night.
Only after cycling through all three of these non-REM stages does your brain enter REM for the first time. That initial REM period is short, typically around 10 minutes. As the night goes on, the balance shifts: deep sleep shrinks in each successive cycle while REM periods grow longer. By the final cycles before morning, a single REM period can last 30 to 60 minutes. This is why you’re more likely to wake up from a vivid dream in the early morning hours than in the middle of the night.
REM Timing Changes With Age
Babies don’t follow the same pattern adults do. Infants under one year old frequently skip non-REM sleep entirely at the start, transitioning directly from wakefulness into REM (called “active sleep” at that age). This direct wake-to-REM pattern gradually fades between 3 and 9 months, but REM latency remains very short throughout infancy compared to older age groups.
Interestingly, healthy children actually have longer REM latencies than adults, meaning they spend more time in non-REM sleep before their first REM period begins. This likely reflects the greater demand for deep, restorative sleep during periods of rapid physical growth. By adulthood, the 60 to 120 minute window stabilizes as the norm.
Why You Might Reach REM Faster or Slower
Several common factors push your first REM period earlier or later in the night.
Sleep Deprivation
When you’ve been short on sleep, your brain compensates with what’s known as REM rebound. Your REM latency can drop below 25 minutes, sometimes dramatically. Even a single one-hour interruption during a normal night of sleep has been shown to trigger unusually fast REM onset when the person falls back asleep. If you’ve ever had an intense, vivid dream during a nap after a rough night, REM rebound is the likely explanation. Your brain prioritizes the REM sleep it missed.
Alcohol
Alcohol delays REM onset and reduces total REM time. Even a low dose, roughly two standard drinks, is enough to disrupt REM sleep. The effect worsens with higher doses in a clear dose-response pattern. This is one reason people who drink before bed often sleep for a full night but wake up feeling unrested. The sleep they got was disproportionately non-REM, skewing the normal balance.
Antidepressants
Several common classes of antidepressants suppress REM sleep, including SSRIs, SNRIs, and tricyclic antidepressants. These medications can extend REM latency significantly. The effect is strong enough that sleep clinics ask patients to stop taking these medications two weeks before diagnostic sleep testing (six weeks for certain longer-acting formulations) so the results aren’t skewed. If you take one of these medications and notice fewer vivid dreams, REM suppression is a likely factor.
When REM Comes Too Soon
Entering REM sleep in under 15 minutes is clinically significant. These episodes are called sleep-onset REM periods, or SOREMPs, and they’re one of the key markers used to diagnose narcolepsy. In people with narcolepsy, the normal gating mechanism that keeps REM sleep at bay during the first cycle breaks down, allowing the brain to slip into REM almost immediately after falling asleep, or even during brief daytime naps.
SOREMPs can also occur in people without narcolepsy under specific conditions: severe sleep deprivation, irregular sleep schedules like shift work, or free-running sleep patterns where a person has no fixed bedtime. Short sleepers, people who naturally need less total sleep, also show a higher tendency toward these rapid REM onsets. The distinction between a normal REM rebound and a clinical problem usually comes down to how frequently it happens and whether it occurs alongside other symptoms like excessive daytime sleepiness or sudden muscle weakness.
How to Track Your Own REM Timing
Consumer sleep trackers (wrist-worn devices and smart rings) estimate sleep stages using heart rate and movement data. They’re not as precise as a clinical sleep study, which measures brain waves directly, but they can give you a general sense of when your first REM period occurs and how much REM you’re getting overall. If your tracker consistently shows your first REM period arriving well before the 60-minute mark or not appearing until very late in the night, that pattern is worth paying attention to.
The most reliable way to support normal REM timing is also the simplest: keep a consistent sleep schedule, limit alcohol close to bedtime, and get enough total sleep so your brain doesn’t need to compensate with rebound. REM sleep is concentrated in the last third of the night, so cutting your sleep short by even an hour disproportionately cuts into REM time rather than deep sleep.