This calculator counts backwards or forwards in whole sleep cycles to suggest bedtimes and wake-up times. The idea is that waking at the end of a cycle, when sleep is lighter, tends to feel easier than being pulled out of deep sleep. The 90-minute cycle it uses is an average, not a measurement of your own sleep, so read the results as a planning aid. It is not medical advice and it cannot diagnose or treat any sleep problem.
What is a sleep cycle?
Sleep is not one block but a repeating sequence of stages. In a sleep study (polysomnography, which records brain waves, eye movement and muscle activity), sleep is divided into non-REM (NREM) and REM sleep. NREM is further split into light sleep (N1 and N2) and deep, slow-wave sleep (N3). You normally drift from N1 to N2 to N3, then back up toward lighter sleep and into REM, and one trip through that sequence is a sleep cycle. In adults a cycle is said to last about 90 minutes on average, commonly reported as somewhere between roughly 70 and 110 minutes, and a night typically contains four to six of them.
Earlier cycles tend to contain more deep sleep (N3), and REM makes up a larger share of the cycles toward morning. That is why cutting a night short is often described as costing you disproportionately much REM sleep. This page sticks to the arithmetic rather than the stage-by-stage detail.
How the calculation works
Bedtime from a wake-up time: bedtime = wake-up time - (number of cycles x cycle length) - time to fall asleep. With the defaults (90 minutes, 15 minutes), six cycles are 540 minutes, so for a 7:00 AM wake-up the answer is 7:00 AM - 9:00 - 0:15 = 9:45 PM the previous evening.Wake-up time from a bedtime: wake-up time = the time you get into bed + time to fall asleep + (number of cycles x cycle length). "Time to fall asleep" means the gap between lying down and actually sleeping. A figure of 10 to 20 minutes is often quoted for healthy adults, but it varies a lot between people, which is why it is an input here and not a constant.
Bedtimes for common wake-up times
Computed when the page is built with 90-minute cycles and 15 minutes to fall asleep. Four cycles are 6 hours of sleep, five are 7.5 hours and six are 9 hours.
| Wake up | 6 cycles (9 h) | 5 cycles (7.5 h) | 4 cycles (6 h) |
|---|---|---|---|
| 5:30 AM | 8:15 PM | 9:45 PM | 11:15 PM |
| 6:00 AM | 8:45 PM | 10:15 PM | 11:45 PM |
| 6:30 AM | 9:15 PM | 10:45 PM | 12:15 AM |
| 7:00 AM | 9:45 PM | 11:15 PM | 12:45 AM |
| 7:30 AM | 10:15 PM | 11:45 PM | 1:15 AM |
| 8:00 AM | 10:45 PM | 12:15 AM | 1:45 AM |
What changes if your cycle is not 90 minutes
The cycle length is the biggest assumption in the whole calculation, and the effect compounds with the number of cycles. The table shows bedtimes for a 7:00 AM wake-up with 15 minutes to fall asleep at four different assumed cycle lengths.
| Cycle length | 4 cycles | 5 cycles | 6 cycles | Sleep for 5 cycles |
|---|---|---|---|---|
| 80 min | 1:25 AM | 12:05 AM | 10:45 PM | 6.7 h |
| 90 min | 12:45 AM | 11:15 PM | 9:45 PM | 7.5 h |
| 100 min | 12:05 AM | 10:25 PM | 8:45 PM | 8.3 h |
| 110 min | 11:25 PM | 9:35 PM | 7:45 PM | 9.2 h |
With 5 cycles, a shift from 90 to 100 minutes per cycle moves your bedtime by 50 minutes. That is a good reason not to treat the displayed time as exact. A window of about 15 minutes either side is a more honest way to read it.
Worked examples
Example 1. You must be up at 7:00 AM and need 15 minutes to fall asleep. Five cycles (450 minutes) means getting into bed at 7:00 AM - 7 h 30 min - 15 min = 11:15 PM. Four cycles would be 12:45 AM and six cycles 9:45 PM. If 11:15 PM is not possible, a time close to 11:30 PM that gets you nearer to five cycles is generally a better target than 12:45 AM, which gives only six hours.
Example 2. You get into bed at 11:30 PM and fall asleep after 15 minutes, so sleep starts at 11:45 PM. Five cycles end at 11:45 PM + 7 h 30 min = 7:15 AM, and four cycles at 5:45 AM. If your alarm is set for 7:00 AM you fall 15 minutes short of the fifth cycle, so an alarm window around 6:45 to 7:15 AM would be about equivalent.
Example 3. A 6:30 AM wake-up with 30 minutes to fall asleep: five cycles give 6:30 AM - 7 h 30 min - 30 min = 10:30 PM, six cycles give 9:00 PM. Those 30 minutes of lying awake cost you a third of a cycle, so being honest about that field matters more than the exact wake-up time.
Limits: do not over-trust the 90-minute rule
- Cycle length differs between people and between nights. Even if the average is about 90 minutes, individual cycles vary, and the first cycle of the night can differ from the later ones. That is why this tool lets you set a value from 70 to 120 minutes.
- Waking at the end of a cycle is not guaranteed to feel better. The explanation that surfacing from light sleep is easier than from deep sleep is intuitive, but a clock-based method like this one cannot tell which stage you are in, and there is no firm evidence that counting cycles on a clock improves how you feel. Use the times as a rough schedule.
- Total sleep matters more than lining up with cycles. A joint consensus statement from the American Academy of Sleep Medicine and the Sleep Research Society (2015) is generally reported as recommending at least seven hours a night for adults. Hitting four neat cycles (six hours) does not make up for too little sleep.
- Grogginess after waking (sleep inertia) can occur even when you wake from light sleep. It usually fades within a short time but varies from person to person.
- Alcohol, caffeine, some medications and sleep apnea can disrupt the structure of sleep itself and make cycle arithmetic meaningless.
- Wearables and sleep apps estimate sleep stages less accurately than a sleep study in a clinic, so treat their stage breakdowns as indicative only.
How to use it in practice
- Fix your wake-up time first. Keeping it within an hour of the same time on weekends helps keep your body clock steady.
- Aim for the five-cycle (or, if that suits you, six-cycle) bedtime, but treat it as the time you start winding down rather than the time the light goes out. Dim lights and reduce screens for 30 to 60 minutes beforehand.
- For a few days, note how rested you feel and how sleepy you get during the day. Then adjust the cycle length between 70 and 120 minutes until the suggested times match the mornings that went well.
- If the issue is the amount you sleep, not the timing, see how late caffeine may be affecting you with the Caffeine Half-Life Calculator, and check how much of your waking day screens take with the Screen Time Life Calculator.
Four common mistakes
- Using the time you get into bed as the time you fall asleep. If you scroll on your phone for 30 minutes before sleeping, five cycles of sleep shrink to about four and a half. Enter your honest time to fall asleep.
- Paying off weekday sleep debt with a huge weekend lie-in. Sleeping three or four extra hours on weekends makes it harder to fall asleep on Sunday night (the shift is sometimes called social jet lag). A smaller difference of one to two hours is easier on the body clock.
- Snoozing the alarm several times. Short stretches of dozing between alarms start a new cycle only to cut it off, which is often blamed for feeling groggier. Setting one wake-up time and getting up, ideally into daylight, is usually better.
- A drink or late coffee close to bedtime. Alcohol can make you drop off quickly yet is reported to make the second half of the night lighter and more broken. Caffeine can stay in the body for several hours, so it is safer to avoid it from the late afternoon on.
Sleep changes with age
Newborns and infants have shorter cycles than adults (often described as roughly 50 to 60 minutes), spend more of their sleep in REM and sleep in several stretches across the day. With age, the share of deep sleep (N3) generally shrinks and night-time waking becomes more frequent. Applying the same 90-minute model to babies, toddlers or older adults is likely to be off, so this calculator is designed for healthy adults.
When not to use this tool
It is a poor fit if you work rotating shifts and your sleep window keeps moving, for infants (whose cycles are shorter), for teenagers and older adults (whose sleep structure differs), and if you suspect insomnia, sleep apnea or restless legs syndrome. If severe daytime sleepiness lasts for a month or more, or you or a partner notice loud snoring with pauses in breathing, see a doctor or a sleep clinic before reaching for a calculator. This page is general information, not a medical diagnosis, and it cannot replace advice from a health professional.
Frequently asked questions
Why does it only show three to six cycles?
Three cycles (4.5 hours) is very short for a night of sleep and is really an emergency option, so four or more cycles are the realistic candidates. Many adults aim for five or six cycles (7.5 to 9 hours).
What happens if I use a value other than 90 minutes?
Ninety minutes is an average, not a rule. Changing the cycle length between 70 and 120 minutes moves the suggested bedtime by several minutes per cycle, as the table above shows. Record a few mornings and adjust the value to match the days when you woke up feeling good.
How do I plan a nap?
Naps are structured differently from night sleep and are outside what this calculator models. Short naps of around 20 to 30 minutes, which usually end before deep sleep, are commonly recommended. A 90-minute nap is sometimes suggested as one full cycle, but a late nap can make it harder to fall asleep at night, so avoid napping late in the afternoon or evening.
Do I need to set my alarm for exactly the calculated time?
No. Because individual cycle lengths vary, allow around 15 minutes of leeway, and remember that keeping one consistent wake-up time matters more than finding the perfect minute.
What if it takes me more than 30 minutes to fall asleep?
If that happens often, look at the likely causes first, such as caffeine, late exercise, screen use and stress, before adjusting the arithmetic. If it lasts two or three weeks or more, consider seeing a sleep specialist. In the meantime, enter the real number in the calculator and your suggested bedtime moves earlier accordingly.