Sleep Calculatorand Scientific Sleep Guides
What time do you need to wake up?
Go to bed at one of these times
Longer and shorter nights
Aim for a 20-minute window around each time rather than the exact minute — cycle length varies from night to night, so the precision here is presentational, not biological. Why that matters →
Adjust the assumptions
The 90-minute default is an average, not a constant. Measured cycles run roughly 70–120 minutes and vary within a single night. If you have tracker data of your own, put your real number in. The evidence on cycle length →
Tell it when you need to be awake. It counts backwards in sleep cycles and gives you the times to aim for — and, unlike every other one of these, it tells you how much confidence that number deserves.
adjustable 70–120
adjustable
AASM guidance
How this calculator works
Four assumptions. All of them stated.
Every sleep calculator on the internet runs the same three lines of arithmetic. Almost none of them tell you what they assumed to get there. Here is the whole thing, so you can decide for yourself whether the answer is worth anything — and here is why we bother doing it this way.
A sleep cycle averages 90 minutes
This is the number everyone uses, and it is a reasonable average. It is not a constant. Individual cycles run roughly 70 to 120 minutes, they vary within a single night, and the first cycle of the night is usually the shortest. One retrospective analysis of 16,441 nights put the median closer to 110 minutes. The slider above exists because your number is probably not 90.
You take about 15 minutes to fall asleep
A systematic review of 110 cohorts covering 4,058 healthy adults put mean sleep onset latency at about 11.7 minutes. Ten to twenty minutes is the normal band. We default to 15 and let you change it, because if you reliably lie awake for 45 minutes, every bedtime below is 30 minutes wrong for you.
Five to six cycles is a full night
At 90 minutes, five cycles is 7 hours 30 minutes and six is 9 hours. Both sit inside the range the American Academy of Sleep Medicine recommends for adults, which is 7 hours or more on a regular basis. Four cycles is 6 hours, which is below it — we still show it, flagged, because sometimes 6 hours is the honest best available.
Waking between cycles feels better than waking mid-cycle
This is the shakiest one, and it is the one every competitor states as fact. Waking out of deep slow-wave sleep does produce worse sleep inertia, so the mechanism is real. But whether counting 90-minute blocks from your bedtime actually lands you in light sleep is a different claim, and it has not been demonstrated. Treat the output as a sensible schedule, not a guarantee about how you will feel.
The arithmetic, in full: bedtime = wake-up time − (number of cycles × cycle length) − minutes to fall asleep. That is it. There is no proprietary model here and nobody else has one either. It all runs inside your browser, so the times you type are never sent anywhere.
Reference
Bedtimes for common wake-up times
Calculated at the defaults: a 90-minute cycle and 15 minutes to fall asleep. If you just want the number without touching anything, it is in this table.
| Wake up at | Bed at (6 cycles · 9h) | Bed at (5 cycles · 7h 30m) |
|---|---|---|
| 5:00 AM | 7:45 PM | 9:15 PM |
| 5:30 AM | 8:15 PM | 9:45 PM |
| 6:00 AM | 8:45 PM | 10:15 PM |
| 6:30 AM | 9:15 PM | 10:45 PM |
| 7:00 AM | 9:45 PM | 11:15 PM |
| 7:30 AM | 10:15 PM | 11:45 PM |
| 8:00 AM | 10:45 PM | 12:15 AM |
| 8:30 AM | 11:15 PM | 12:45 AM |
| 9:00 AM | 11:45 PM | 1:15 AM |
Read this before you trust the minutes
These times are precise to the minute because arithmetic is precise to the minute. Your sleep is not. Aim for a 20-minute window around whichever row applies to you, and judge the schedule over a fortnight rather than over one morning.
Bigger levers
Three things that matter more than cycle timing
If you optimise the minute you go to bed and ignore these, you have tuned the smallest dial on the machine.
Keeping the same schedule
A UK Biobank analysis of 60,977 people, using over 10 million hours of accelerometer data, found that how regular someone's sleep timing was predicted all-cause mortality more strongly than how long they slept. Going to bed at a consistent time beats going to bed at the mathematically optimal time.
Getting enough hours at all
Cycle alignment is a rounding error next to total sleep. Five perfectly-timed cycles is still 7h 30m; four perfectly-timed cycles is 6 hours and no amount of timing rescues it. Get the duration into range first, then worry about where the boundaries fall. What your range is →
Morning light and what you do on waking
Work on 833 twins and unrelated adults linked how alert people felt after waking to sleep the night before, physical activity the previous day, and breakfast composition — not to genetics. Sleep inertia typically clears in 15 to 30 minutes regardless. The morning is more adjustable than the arithmetic suggests.
Questions people actually ask
Straight answers, including the unhelpful ones
Go deeper
The reading behind the numbers
What a sleep cycle actually is
Stages, proportions, why deep sleep is front-loaded, and where the 90-minute figure came from.
How much sleep do I actually need
The AASM and NSF numbers by age, what “need” means, and why your personal number may sit outside the band.
Are sleep calculators accurate?
I built one. Here is exactly where it is right, where it is guessing, and what I use it for myself.
Same time every night
If you take one thing from this page, take that one. The calculator gives you a time. The benefit comes from hitting it on a Tuesday and on a Saturday.
Read the field notes