How long should a power nap be? What NASA, and a 10-vs-30-minute trial, found

The two studies everyone quotes

The first is NASA’s. In the 1990s the agency’s fatigue countermeasures group let long-haul pilots take a planned rest in the cockpit seat during the cruise phase of real transpacific flights. The pilots who rested slept for an average of 26 minutes; compared with a no-rest group they were 34% better on a reaction-time task and 54% more alert on physiological measures later in the flight. That is where the “26-minute nap” comes from. It was a rest window of about 40 minutes, with the sleep inside it averaging 26.

The second is a 2006 laboratory comparison by Brooks and Lack. After a night of restricted sleep, volunteers took naps of 5, 10, 20 or 30 minutes — or none — in the early afternoon, and were tested for three hours afterwards. The 10-minute nap gave immediate improvements in alertness and performance that lasted up to two and a half hours, with no grogginess. The 20-minute nap helped too, but its benefits took about 35 minutes to appear. The 30-minute nap produced a period of impaired alertness immediately on waking before it paid off. Five minutes did little.

Why 30 minutes turns a corner

The explanation is sleep stages. For the first ten to twenty minutes you are in light sleep. Somewhere around the 25–30 minute mark many people drop into slow-wave sleep — the deep stage — and waking from it produces sleep inertia: the heavy, slow, almost drunk feeling that can last from a few minutes to half an hour. A short nap never gets there, which is the whole trick. A long nap of about 90 minutes goes all the way through deep sleep and REM and tends to end in light sleep again, which is why a full cycle can also feel clean — it just costs an hour and a half and more of tonight’s sleep pressure.

The nap-length pages go through each duration from 10 to 90 minutes.

The clock, in practice

Add fifteen minutes to any nap for falling asleep. A “20-minute nap” is a 35-minute alarm. Lying down at 1:30 pm, that is an alarm at 2:05; a 26-minute NASA-style nap means 2:11; a full cycle means 3:15.

Timing matters as much as length. Early afternoon suits most people: alertness naturally dips after lunch, and a 1:30 pm nap leaves more than eight hours before a 10:45 pm bedtime. Later than about 3 pm, a nap that reaches deep sleep starts to take pressure away from the night — the same mechanism that makes a late coffee a problem. The nap calculator shows the gap to your bedtime for any start time.

What a nap cannot do

It cannot replace the night. The Brooks and Lack volunteers were sleep-restricted, and the nap helped them get through an afternoon; it did not repay the debt. If you need a nap most days to function, the number to look at is tonight’s bedtime, not the afternoon’s. The sleep-needs pages give the range for your age, and the bedtime pages turn your alarm into a lights-out time.

Questions people ask

Is a 30-minute nap too long?

It is the edge. In the Brooks and Lack comparison, 30 minutes produced a period of impaired alertness on waking before the benefits appeared. Twenty minutes avoided that.

What time of day is best for a nap?

Early afternoon, roughly 1 to 3 pm: it matches the natural dip in alertness and leaves eight-plus hours before a normal bedtime.

Does a coffee before a nap work?

The idea is that caffeine takes about 20 minutes to act, so it kicks in as a short nap ends. Small studies support it for alertness; it is not useful if you then need to sleep well that night.