Nap calculator: when to set your alarm

Work out your alarm time

Set the time you are lying down. The calculator adds an allowance for falling asleep, then gives you the alarm time for each nap length. The default of 12 minutes is the median from my own 28-night diary rather than a round number, but change it to yours if you know it.

What each nap length actually does

These are not rules of thumb. They come from the one experiment that put nap durations side by side: Brooks and Lack (2006), 24 healthy adults, each taking all five conditions after a night cut to five hours, tested repeatedly for three hours afterwards.

The 60 to 90 minute row comes from somewhere else: Mednick, Nakayama and Stickgold in Nature Neuroscience (2003), out of Harvard, who found that naps of that length containing both slow-wave and REM sleep produced learning gains on a perceptual task comparable to a full night of sleep. That is memory consolidation. It is a different goal from feeling awake this afternoon, and the two should not be treated as settings on one dial.

The zone to avoid, and why it exists

Slow-wave sleep, stage N3, typically does not begin until roughly 30 minutes into a nap. Wake out of N3 and you get sleep inertia: genuinely impaired alertness and cognition that can take 15 to 30 minutes to clear. It shows up on reaction-time testing, so it is not simply a feeling of grogginess.

That is the whole logic behind the standard advice. Stay under about 30 minutes and you should still be in lighter N1 and N2 sleep when the alarm goes, and wake cleanly. Go past 90 and you have completed a cycle and are surfacing naturally anyway. The trouble is the middle.

Adding your falling-asleep time is the part most nap calculators skip. If you set a 30-minute timer and take 10 minutes to drop off, you have booked yourself a 20-minute nap. If you take 2 minutes, you have booked a 28-minute one and may wake up in deep sleep. The calculator above asks for that number rather than assuming it.

When to nap

Early afternoon, roughly between 1pm and 3pm, which lines up with the natural dip in alertness most people feel after lunch. Napping in that window tends to feel restorative rather than disorienting.

The later constraint matters more than the earlier one. Sleep pressure builds through the day and a nap spends some of it, so a nap close to bedtime can make falling asleep at night harder. The calculator flags this if you give it your bedtime. Six hours is the gap it uses, and I want to be clear that this is a practical rule of thumb rather than a threshold established by a trial.

What my own falling-asleep time looks like

The default above is 12 minutes rather than a round 10, and it comes from data rather than from a textbook range. I kept a nightly diary for the 28 nights of February 2026, the 1st to the 28th, recording lights-out at the time, an estimate the following morning of how long I took to fall asleep, and a handful of things that might have affected it. Twenty-eight consecutive nights, no nights dropped, including the bad ones.

In a 28-night self-tracking diary run from 1 to 28 February 2026, estimated sleep latency had a median of 12 minutes, with 50% of observations between 7 and 19 minutes.

The spread is the useful part. A single default of 12 minutes is right for the middle of my own distribution and wrong for a fifth of my nights. On a bad night I would have needed to add half an hour, not twelve minutes, and a nap alarm set on the assumption of twelve would have put me straight into the window this page tells you to avoid.

Which is the honest argument for the input box rather than a fixed number: your median is not mine, and your own bad nights are not your median either.

What moved it, and what that does not prove

I logged five things alongside the estimate. The differences between subgroups were large:

I want to be careful about what this can and cannot show, because the temptation is to read it as six independent effects and it is nothing of the kind.

• These factors travel together. A high-stress night is also more likely to be a night with late caffeine, a screen before bed, and no exercise. The diary cannot separate them, so the rows are not additive and none of them is isolated.
• The subgroups are small. Twenty-eight nights split six ways leaves very few observations in each cell. A difference of several minutes between two small medians is not a finding.
• This is one person. It says something about me and nothing about you, which is exactly why the calculator asks rather than assumes.
• Latency here is estimated, not measured. Self-reported sleep onset is known to be unreliable, and people generally overestimate it. Measuring it properly needs equipment I do not have.

What I would actually take from it: the two largest gaps are stress and late caffeine, and they are the two I would look at first if my own falling-asleep time started drifting. That is a hypothesis worth testing, not a result.

The one I found most interesting, and trust least

I also rated how rested I felt each morning, out of five.

It is a clean-looking gradient and I do not think it means what it appears to mean. The obvious reading is that falling asleep faster produces better mornings. The more likely reading is that a third thing, stress most obviously, produces both the long latency and the bad morning, and the two are consequences of each other rather than one causing the other. Long latency also costs you sleep if the wake time is fixed, which would produce the same pattern without any direct effect at all.

So: a real association in my own data, three plausible explanations, and no way to tell them apart with a diary. I am reporting it because leaving out the result I like most would be the wrong instinct for this site, not because it demonstrates anything.

What this calculator assumes

Every tool like this makes assumptions. Most do not print them. Here are mine.

• That you fall asleep in the time you told it. The default of 12 minutes is the median from my own 28-night sleep diary, not a textbook figure. It varies enormously: my own 90th percentile was 31 minutes, and in the NASA cockpit study fatigued pilots took about 5.6 minutes.
• That deep sleep starts around the 30-minute mark. An average. Yours may be earlier or later, particularly if you are short on sleep.
• That the Brooks and Lack findings transfer to you. Their participants were 24 healthy young adults, all good sleepers, all following a night restricted to five hours. That is napping while sleep-deprived, which is not the same as napping on a normal day.
• That six hours before bed is a sensible cutoff. A rule of thumb, not a finding.

If any of those does not describe you, the times above are a starting point rather than an answer. That is true of every nap calculator on the internet. This one just says so.

Questions people actually ask

Related reading

• How long should a nap be? The 20-minute rule, traced, where the 20-minute figure came from, what the NASA study actually measured, and why the number everyone repeats is a ceiling rather than an optimum.
• Where sleep numbers come from, the source index: the 20-minute nap alongside seven other sleep claims traced to their origins.
• How many hours of sleep is 11pm to 7am?, for the night rather than the afternoon.
• How long a sleep cycle actually is, why the 90 in "90-minute nap" is a midpoint of 70 to 120.