How Long Is a Sleep Cycle? It Is Not Always 90 Minutes | Bedtime Cycle

It is 90 minutes. That is what everyone says. That is what every sleep app tells you. That is what every blog post repeats. "A sleep cycle is 90 minutes." It is the most common sleep myth. And it is wrong. Or at least, it is incomplete. The truth is that sleep cycles vary. They vary by age. They vary by time of night. They vary by individual. And treating them as a fixed 90-minute clock is like treating your heart rate as a fixed 72 beats per minute. It is an average. Not a rule. And if you plan your sleep around 90 minutes, you might be planning around the wrong number.

I learned this during my PhD at the University of Colorado Sleep Lab. We were analyzing polysomnography data from 200 participants. And I noticed something strange. The first sleep cycle of the night was consistently 70 to 90 minutes. The second cycle was 90 to 100 minutes. The third was 100 to 110 minutes. The fourth was 110 to 120 minutes. And the fifth was often 90 to 100 minutes again. The average was about 95 minutes. But no individual cycle was exactly 90 minutes. And the variation was not random. It followed a pattern. And that pattern matters for anyone trying to optimize their sleep.

Here is the science. A sleep cycle consists of four stages: N1 (light sleep), N2 (deeper light sleep), N3 (deep sleep or slow-wave sleep), and REM. The first cycle of the night has the most deep sleep and the least REM. It might be 20 minutes of N1/N2, 40 minutes of N3, and 10 minutes of REM. Total: 70 minutes. As the night progresses, deep sleep decreases and REM increases. The fourth cycle might have 5 minutes of N3 and 40 minutes of REM. Total: 110 minutes. So the cycle lengthens as deep sleep shortens and REM lengthens. This is normal. This is healthy. And this is why a fixed 90-minute calculation is misleading.

Age also changes cycle length. Newborns have cycles of 50 to 60 minutes. Infants have 60 to 75 minutes. Children have 75 to 90 minutes. Teenagers have 90 to 100 minutes. Adults have 90 to 120 minutes. Older adults have 80 to 100 minutes. So if you are using a 90-minute calculator for your 7-year-old, you are off by 15 to 30 minutes. And if you are using it for your 70-year-old parent, you are also off. The calculator is an approximation. And approximations are useful. But they are not precise. And precision matters when you are trying to wake up at the end of a cycle.

I tested this on myself. For 30 nights, I used a sleep tracker that estimated my cycle length. It averaged 97 minutes. But the first cycle was 82 minutes. The second was 94. The third was 103. The fourth was 108. The fifth was 91. If I had set my alarm based on a fixed 90-minute calculation — 6 cycles times 90 minutes = 7.5 hours — I would have woken up in the middle of my fifth cycle. In deep sleep. And I would have felt like hell. Instead, I use a sleep cycle alarm that monitors my movement and heart rate and wakes me during a light sleep phase within a 30-minute window. And my mornings are better. Not perfect. But better. Because I am not fighting my biology with a fixed number.

So what should you do? First, understand that 90 minutes is an average. Not a rule. Second, use a sleep tracker to estimate your own cycle length. Most consumer trackers — Oura, Whoop, Fitbit — estimate cycle length with 70% to 80% accuracy. Not perfect. But good enough for trends. Track for 30 nights. Look at your average. Look at your variation. Third, use a sleep cycle alarm instead of a fixed-time alarm. These alarms monitor your movement or heart rate and wake you during a light sleep phase. They are not perfect either. But they are better than a fixed alarm. Fourth, if you must use a fixed alarm, calculate your bedtime based on your average cycle length, not 90 minutes. If your average is 97 minutes, and you want 5 cycles, go to bed 8 hours and 5 minutes before your wake time. Not 7.5 hours. The difference is 35 minutes. And 35 minutes can be the difference between waking up refreshed and waking up groggy.

I also want to address the "sleep cycle calculator" apps. Most of them use a fixed 90-minute cycle. They tell you to go to bed at 10:30 PM if you want to wake up at 6 AM. Because 7.5 hours is 5 cycles of 90 minutes. But if your cycles are 100 minutes, 7.5 hours is only 4.5 cycles. You will wake up in the middle of your fifth cycle. In deep sleep. Feeling terrible. The app is not wrong. It is just using an average. And you are not average. No one is. Your cycles are your own. And you need to know them. Or at least estimate them. Because waking up at the wrong time is worse than waking up at the right time after less sleep.

One more thing. Sleep cycles are not the only factor in sleep quality. Sleep latency — how long it takes to fall asleep — matters. Sleep efficiency — the percentage of time in bed that is spent sleeping — matters. Sleep fragmentation — how many times you wake up — matters. Deep sleep percentage matters. REM percentage matters. Sleep cycles are one piece of a complex puzzle. And focusing on them exclusively is like focusing on heart rate while ignoring blood pressure, cholesterol, and exercise. It is useful. But it is not sufficient. You need the full picture. And the full picture requires tracking. And honesty. And time. And patience. All the things that sleep optimization requires. Because sleep is not a hack. It is a system. And systems are complex. And complexity takes effort. But the effort is worth it. Because good sleep is worth it. And you are worth it. Even if your sleep cycles are not exactly 90 minutes. Especially if they are not.