Naps: The 10-Minute Rule and When Not to Use It

Napping advice usually comes as a vibe: short naps good, long naps bad, something about 90-minute cycles. The vibe happens to be roughly right, which is rare, but the actual numbers are more specific and more useful than the folklore. There is one small trial that did the obvious experiment properly, and its result is the closest thing napping has to a rule.

What is the best nap length?

Brooks and Lack took 24 healthy adults, restricted them to 5 hours of night sleep, and gave them naps containing exactly 5, 10, 20 or 30 minutes of sleep, measured by EEG rather than time on the mattress. Then they tracked alertness, fatigue and cognitive performance for three hours.

The 10-minute nap won on nearly everything: improvements across sleepiness, fatigue, vigour and performance, appearing immediately and lasting around 155 minutes. The 5-minute nap did almost nothing. The 20-minute nap helped, but its benefits took half an hour or more to arrive. The 30-minute nap caused a period of impaired alertness right after waking before its benefits showed up at all.

One trial, 24 people, sleep-restricted young adults. I would not carve it in stone. But it is the cleanest dose-response data napping has, and its shape matches both the mechanism below and what most nappers eventually learn by trial and error.

Why longer naps make you feel worse

Sleep is not a uniform substance you collect by the minute. Around the 20 to 30-minute mark, a napping brain starts descending into slow-wave sleep, and being woken from slow-wave sleep produces sleep inertia: the heavy, disoriented, worse-than-before state that can drag on for half an hour or more.

This is the entire logic of the short nap. It is not that 10 minutes of sleep is magically restorative; it is that 10 minutes ends before the trapdoor opens. A 90-minute nap avoids the trapdoor from the other side, by completing a full cycle, but at the price of discharging so much sleep pressure that the coming night pays for it.

Note that the clock starts at sleep onset, not when you lie down. If you take 10 minutes to drift off, a 20-minute alarm is really a 10-minute nap. That detail is why the trial measured EEG sleep, and why your alarm needs a small allowance for your usual latency.

When should you nap?

The window is early afternoon, roughly 1 to 3 pm, for two reasons that point the same way. The circadian dip after lunch makes falling asleep for a nap actually feasible. And a nap taken there still leaves seven or more hours for sleep pressure, the accumulated adenosine that makes you sleepy, to rebuild before a 10:30 pm bedtime.

A nap at 5 pm sits too close to the night it borrows from. This is the same accounting as the caffeine cutoff: both an afternoon espresso and an evening nap steal from the night, one by blocking adenosine, the other by spending it.

The coffee nap is not a joke

Drink a coffee, then immediately take a 15 to 20-minute nap. It sounds like a bit, but Reyner and Horne tested it in sleepy drivers in a simulator and the combination suppressed driving impairment better than either intervention alone. The mechanics line up neatly: caffeine needs about 20 to 30 minutes to reach peak blood levels, so it arrives exactly as the alarm goes off, on a brain the nap has just partially cleared of adenosine.

I would reserve it for genuine emergencies, a long drive or a wrecked night before a demanding day, rather than daily use, because the caffeine half of the trick still counts against your cutoff time. As emergencies go, it is the best-tested one available.

Who should not nap at all

If you have trouble falling asleep at night, naps are working against you. Sleep pressure is the main force that puts you under at 11 pm, and a nap spends it early. This is why behavioural insomnia treatment, the CBT-I mentioned in sleep hygiene, typically bans daytime sleep outright: the deficit is being deliberately concentrated into the night.

The same logic runs in reverse for shift workers and new parents, where the night is structurally broken and a nap is patching a real hole rather than draining a working system. Rules follow the situation. A nap is a tool for people whose nights work; it is a leak for people whose nights do not.

Do long naps mean something is wrong?

Cohort studies keep finding that habitual long naps, an hour or more, associate with worse cardiovascular outcomes and higher mortality. I read most of that as reverse causation: undiagnosed sleep apnoea, depression and chronic illness all produce long daytime sleep, and cohort adjustments cannot fully unpick it. A deliberate 15-minute nap and a helpless daily 90-minute crash are different phenomena that happen to share a word.

The useful signal is the second one. If you need long naps to function despite spending 8 hours in bed at night, that is not a napping preference, it is a symptom, and the screening list in sleep and longevity applies.

A napping decision rule

Everything above compresses into four questions, answered in order:

1. Do you take more than 30 minutes to fall asleep at night, or wake often? Then no naps at all until the nights work. Fix the night first; the afternoon slump is a symptom, not a separate problem.
2. Is it after 3 pm? Then no nap today. Coffee is also off the table at this hour, so the honest options are a walk, daylight, or accepting a mediocre afternoon.
3. Is this a normal day? Alarm at 20 minutes, done. Judge the result at 4 pm, not at the moment of waking.
4. Is this an emergency, a long drive, a red-eye, an exam? Coffee nap: caffeine first, 20-minute alarm, and count the caffeine against tomorrow’s cutoff.

The rule I break most often is the first one. After a bad night the nap is exactly the thing that feels most justified, and taking it is how one bad night becomes a bad week. If the night was short for a one-off reason, a 10-minute nap is fine. If short nights are the pattern, the nap is feeding the pattern.

What we do not know

The 10-minute result comes from one lab under sleep restriction; nobody has replicated it at scale or in well-rested adults, where the appetitive nap may behave differently. Habitual nappers seem to extract more benefit than novices and we do not fully know why. And the long-nap epidemiology cannot currently separate naps that cause problems from naps that reveal them.

Where to start

If your nights are fine and your afternoons sag: alarm set for 20 minutes, sometime between 1 and 3 pm, lying down, no phone. Judge it by the two hours that follow, not by how the waking moment feels. If your nights are not fine, skip this article’s advice entirely and run the sleep protocol first; the nap question answers itself once the nights work.