90 minute sleep cycle calculator

The 90-minute rule, applied properly: for a 6:30 am alarm, five 90-minute cycles plus 15 minutes to fall asleep puts you in bed by 10:45 pm. This calculator does that arithmetic with your wake time — and lets you move the cycle length off 90, because it was always an average, never a constant.

Direction
90 min

90 min is the average — yours may differ.

15 min

10–20 min is typical for a healthy sleeper.

Go to bed at one of these times

121518210003060912bed 10:45 pmwake 6:30 amREM share of each cycle — it builds toward morning, so the last cycles are the ones to protect

the arithmetic · 5 cycles × 90 min + 15 m to fall asleep = 7 h 45 m in bed → in bed by 10:45 pm to wake at 6:30 am

An alarm near a cycle boundary may reduce sleep inertia for some people, but cycle length varies from night to night. Treat these as planning estimates. The highlighted option gives most adults enough total sleep; pick a longer one if you're repaying debt.

NextBuild tonight's wind-down routine around this 10:45 pm bedtime — caffeine cutoff, dim lights, and screens off, each step timed backwards. Your selected bedtime carries into the planner.

Planning tool, not medical advice. These figures are population averages for scheduling your sleep — they don't diagnose or treat any condition. See a clinician for persistent sleep problems.

What is the 90-minute sleep cycle rule?

Sleep is not a flat state: the brain loops through light sleep, deep slow-wave sleep and REM in cycles, and in laboratory recordings one loop averages about 90 minutes. The practical rule that fell out of that finding: waking is easiest at the seam between two cycles, so bedtimes and alarms should be planned in 90-minute blocks rather than round hours. It is a genuinely useful heuristic — as long as you remember it describes an average night for an average sleeper.

Where does the 90-minute rule break down?

In two places. Between people: measured cycle length runs roughly 70–120 minutes, so a hardcoded-90 bedtime can land mid-cycle for you on a predictable schedule. And within one night: the first cycle is often shorter and dominated by deep sleep, while the last runs longer and carries most of the night's REM. The fix is not a better constant — it is treating the 90 as a dial. Calibrate it: hold a fixed wake time for a week, shift the cycle length, and keep the value that has you waking clear-headed. The sleep cycle calculator is the same engine with the full explanation of the model.

Why do most 90-minute calculators get the bedtime wrong?

They count cycles and stop. But cycles measure sleep, and you are not asleep at lights-out — healthy sleep latency is 10–20 minutes. A calculator that ignores it hands you a bedtime that is quietly 15 minutes too late every single night: your first cycle starts while you are still awake, and the whole 90-minute lattice shifts. This one adds the latency first (default 15, adjustable 0–60) and only then counts cycles — the formula is printed with your numbers under the timeline, every time. Pair the result with a fixed wake time from the bedtime calculator and the schedule holds itself.