Sleep CalculatorBedtime, Wake-Up Time & 90-Minute Cycles

How much sleep you need depends on your age more than any other single factor. Public-health guidance from the U.S. Centers for Disease Control and Preventionand the American Academy of Sleep Medicine (Watson et al., Sleep, 2015) converges on the ranges by age band shown below. These are the numbers WorkItOut.Tools' age selector uses as defaults.

Two things this chart doesn't tell you. First, ranges are ranges - individuals within an age band vary by an hour or two in either direction, and the honest test of your personal number is which end of the range leaves you consistently feeling rested in the morning. Second, sleep need shifts with life circumstances: recent sleep debt, illness, pregnancy, elite training loads, and grief all temporarily raise the requirement.

The ranges also blur near their edges. A late-teen night owl and a 20-year-old university student often have almost identical biology, even though they sit on either side of a band boundary - sleep science is fuzzier than any bracket lets on.

Pick your band in the calculator above, then use the "Aiming for" chip on the result to nudge from the midpoint if you know your own number better. The chip covers 5 to 12 hours regardless of age band, so power users aren't boxed in by the recommendations.

Sleep unfolds in recurring cycles through four stages - light non-REM (N1 and N2), deep non-REM (N3), and REM. That much is settled sleep science. What isn't settled is that every cycle lasts exactly 90 minutes.

The 90-minute figure that every viral bedtime calculator uses is an average. Actual cycles range from about 80 to 110 minutes across the night, and they typically lengthen as the night progresses - earlier cycles have proportionally more deep sleep, later ones have proportionally more REM (Sleep Foundation, Stages of Sleep). So aligning your alarm to a cycle boundary calculated as "wake time minus five times 90 minutes" is a rough approximation, not a biological guarantee.

That doesn't make the method useless. Waking during a lighter sleep stage does feel less groggy than waking mid-N3, and cycle math gets you roughly to a lighter-sleep window even when the arithmetic is imprecise. It's a useful frame for choosing between "roughly 7.5 hours" and "roughly 9 hours" - the difference between five and six 90-minute cycles is meaningful even when the boundaries drift.

WorkItOut.Tools defaults to the 90-minute cycle method because that's what people type when they search for "sleep calculator". But the result card shows both a duration answer and the cycle options side by side, with the cycle answer flagged as an approximation. You can compare, then pick.

Sleep latency - the time it takes to actually fall asleep after getting into bed - is the number most sleep calculators quietly ignore. And it's the reason most of them are wrong.

Healthy adults typically take 10 to 20 minutes to fall asleep after settling in (Sleep Foundation, Sleep Latency). Consistently under 5 minutes can signal sleep deprivation; consistently over 30 minutes may point to insomnia. That's a real, measurable delay - but it usually doesn't show up in a sleep calculator's output.

Most calculators hard-code a single latency assumption in the background, if they use one at all. That's often 15 minutes. So when a competitor tells you to "go to bed at10:45 PM" for an 8-hour target and a7 AM wake, they're actually assuming you'll be asleep by11:00 PM. If your true latency is 25 minutes, you're 10 minutes short of your target every night - 70 minutes of sleep debt per week from one silent default.

WorkItOut.Tools makes latency explicit. It's a first-class setting in the Advanced cog, defaulting to 15 minutes but adjustable from 0 to 120. The result card reports both "get into bed" and "estimated asleep" as separate times, so you can see the delay in your plan, not buried in an assumption. If you know you drift off in five minutes or lie awake for forty, tell the calculator - the answer changes.

Why do sleep calculators give different answers? Because they don't agree on what they're calculating. Two families of methods dominate the space, and each answers a different question.

The Duration method calculates from a target number of hours. Feed it your wake time, subtract however many hours you're aiming for, and it hands back a bedtime. This is the approach public-health bodies like theCDC and the American Academy of Sleep Medicine (Watson et al., Sleep, 2015) actually recommend - it's built around the number that matters biologically, total sleep.

The 90-minute cycle method calculates from a target number of complete sleep cycles. It counts backwards from your wake time in 90-minute chunks and rounds to the nearest cycle boundary. It is the method that goes viral on TikTok, and it is rooted in real sleep science - cycles are a thing - but the 90-minute figure is an average, not a fixed clock.

Both methods produce different answers for the same input. Neither is wrong; they optimise for different things. Most sleep calculators pick one method and hide the other, which is why two calculators can give you 45-minute-apart bedtimes for the same 7 AM wake-up.

WorkItOut.Tools shows both. Pick the method that suits your goal, or run the numbers through both and see how far apart they land. Every assumption is visible - no hidden defaults, no invisible math.