How long should a nap be, and when should you skip it?

It is quarter past two. The meeting after lunch took the last of what your brain had, and you have now read the same paragraph three times. Somewhere at the back of your mind sits the thought of twenty minutes in the car, or on the sofa, or face down on the kitchen table.

It is a reasonable thought. The nap is one of the very few tools that genuinely does something about tiredness in a couple of minutes. It is also one of the most misunderstood things on the whole health shelf, because almost every piece of advice about it leaves out the part that decides whether it helps or hurts: what it costs you tonight.

This guide is about daytime sleep. What makes nights better in general is covered in the guide to sleeping better after 40. Why you wake in the middle of the night is covered in the guide to waking at 3am. How the timing of training affects the night is covered in the guide to exercise and sleep. This one is about those twenty minutes in the middle of the day.

First: the afternoon dip is not lunch's fault

The usual explanation for two o'clock is food. You ate too much, you ate too many carbohydrates, the blood went to your stomach. That explanation is partly wrong.

Timothy Monk summarised the research in a 2005 review that is still cited. His conclusion: the post-lunch dip is a real phenomenon that can occur even when the person has had no lunch at all, and even when they are unaware of the time of day. Its roots are in human biology rather than on the plate. Monk attaches two caveats worth carrying with you. The dip shows up for some performance measures in some individuals, not all measures in everyone. And a high-carbohydrate lunch certainly makes it worse, even if it does not cause it.

The practical consequence: blaming lunch leads to the wrong fix. Eating lighter at noon softens the dip but does not remove it. The dip is a time of day, not a meal.

A nap is a withdrawal, not a free lunch

This is the point of the page, and it explains every rule further down.

The framework sleep research has used since Borbély's original 1982 paper is the two-process model. Borbély and colleagues revisited it in 2016 in the Journal of Sleep Research. In short, your sleep is governed by two things working at once. Process C is the body clock: the internal rhythm that makes you alert or sleepy at particular times regardless of how long you have been awake. Process S is sleep pressure: a pressure that builds through every waking hour and discharges while you sleep.

Sleep pressure is a balance, then. At breakfast it is low. At bedtime it is high, which is the entire reason you fall asleep. A nap in the middle of the day spends part of that balance.

Once you see it that way, both nap rules follow on their own. It should be short, because then the withdrawal is small. It should be early, because then the pressure has time to rebuild before the evening. A long nap at six in the evening breaks both rules at once: a large withdrawal with no time to rebuild. That is why one of those makes bedtime so difficult.

How long? Two very different kinds of evidence point the same way

The first is public health guidance. Sweden's health service 1177 writes that if you need it, you can sleep for a short spell in the middle of the day, a so-called nap or power nap, and that to avoid disturbing the coming night it should be taken before 3pm and last a maximum of 15 minutes.

The second comes from a laboratory in Australia. In 2006, Brooks and Lack had twenty-four participants sleep about five hours at home. They were then given an afternoon nap at 3pm containing exactly five, ten, twenty or thirty minutes of sleep, plus a no-nap control condition, and were tested for three hours afterwards.

The measures were sleep latency, subjective sleepiness, fatigue, vigour and cognitive performance. The authors concluded that ten minutes was the most effective of the durations tested.

Notice what the table actually says. It is not that a longer nap gives you less. Thirty minutes gave just as long a benefit as ten. The difference is in when the benefit arrived. Someone with twenty minutes between two meetings who sleeps thirty gets the impairment and never reaches the improvement.

Two completely different kinds of evidence, a public health recommendation and a laboratory experiment, still land on the same message: keep it short. That is unusual enough to be worth trusting.

Translating that into practice

Allow for the time it takes to drop off. If you need five minutes to fall asleep and you want ten minutes of sleep, the alarm should be set for roughly fifteen to twenty minutes from lying down. Set it before you lie down, never afterwards.

Sleep inertia: why you wake up worse than you went down

The most common reason people dismiss naps as something that does not work for them is sleep inertia. Hilditch and McHill summarised the field in 2019: sleep inertia is the grogginess felt on waking, it is associated with measurable decrements in cognitive performance, and it dissipates as time awake increases. It has been observed both in tightly controlled laboratory studies and in real-world settings.

Two things make it worse, according to the same review. Prior sleep loss. And the time of day at which you wake. Both apply to the person who is napping precisely because the night was short.

Which is why the table above is so useful. Sleep inertia is not a sign that the nap was unnecessary. It is usually a sign that it was too long.

The question most people skip: how do you sleep at night?

Here is the part that separates this page from most power-nap articles. The advice about napping is not one piece of advice. It is two, and which one applies to you is decided by your nights.

Do you sleep broadly well? Then the nap is a tool. Short, early, when needed. Use it on the days it is needed.

Do you have recurring trouble sleeping at night? Then the advice reverses. Under the heading "do not sleep during the day", 1177 writes that many people who sleep badly at night want to sleep in the daytime, but that the coming night is affected if you fall deeply asleep and sleep for a long time. Their suggested alternative for getting more energy is a relaxation exercise.

In 1177's structured sleep programme this is not a footnote but a step in the treatment. It states that anyone who struggles to sleep at night should avoid daytime sleep because it worsens the night, and that people who rest a lot or sleep long during the day get worse nights. The recommendation is to move and stay active during the day even when the tiredness argues otherwise.

The logic is exactly the balance from earlier. If you are short of sleep at night, sleep pressure is the one thing working in your favour. Spending it at 2pm makes the evening harder, which leaves you more tired the next day, which makes the nap more tempting. That is the vicious circle 1177 warns about.

The nine o'clock sofa doze is a nap too

The most common nap taken in most homes is not taken in the middle of the day. It is taken half-sitting on the sofa somewhere between nine and eleven at night with the television on, and it is never called a nap.

Measured as a balance, it is the most expensive sleep you can take. Sleep pressure has built all day, and you spend part of it twenty minutes before you need it. Then you go to bed and wonder why you are lying there awake. Plenty of people who describe themselves as bad at falling asleep in fact fall asleep beautifully. Just on the wrong piece of furniture.

The fix is dull but effective: when your eyes start closing on the sofa, that is the signal to go to bed rather than to stay put. One more episode costs more than it looks like it costs.

Is napping dangerous? What the epidemiology actually shows

Headlines appear at intervals claiming naps are linked to heart disease or early death. The numbers are worth looking at, because they say something rather different from the headlines.

The largest synthesis is a 2020 dose-response meta-analysis by Pan and colleagues covering 313,651 people from 20 cohort studies, of whom 38.9% napped. The findings:

• Napping overall was linked to higher all-cause mortality, hazard ratio 1.19 (95% confidence interval 1.12 to 1.26).
• For incident cardiovascular disease there was no association at all in the pooled analysis.
• In subgroups an association did appear: 1.31 in women, 1.36 in older participants and 1.34 among those taking long naps.
• The dose-response analysis produced a J-shaped curve for cardiovascular disease. Risk fell from zero up to around 25 minutes a day, then rose sharply. For all-cause mortality the relationship was a straight upward line instead.

Note where the bottom of that J sits. Around 25 minutes. In practice that is the same length the laboratory and 1177 arrived at for entirely unrelated reasons.

But it is worth being honest about what this kind of study can and cannot do. It measures associations in people who chose to sleep during the day. Fatigue is a symptom. Illness makes people tired. Sleep apnoea makes people tired. Long, frequent naps being followed by worse health may just as easily mean that worse health produces long, frequent naps. The authors themselves call for larger studies and genetic studies to work out the mechanism.

The closest thing we have to a test of that objection is a Swiss cohort study published in Heart in 2019. Häusler and colleagues followed 3,462 people with no previous cardiovascular disease for just over five years. The outcome:

• People who napped once or twice a week had a lower risk of cardiovascular events, hazard ratio 0.52 (interval 0.28 to 0.95), in both unadjusted and adjusted models.
• People who napped almost daily, six to seven times a week, had an unadjusted raised risk of 1.67. After adjustment for major cardiovascular risk factors, plus excessive daytime sleepiness or obstructive sleep apnoea, it fell to 0.89 (interval 0.58 to 1.38), leaving no raised risk at all.
• No association whatsoever was found between nap duration and cardiovascular events.

That is exactly the pattern you would expect if the original association came from why people sleep rather than from the fact that they sleep. The study is small in this context, though, with 155 events in total, and the interval 0.28 to 0.95 sits close to the line of no effect. It is an argument against the scare headlines rather than an argument that naps protect your heart.

What to take from it: nothing in this evidence argues against a short nap a few times a week. A daily need to sleep at length during the day is a different matter and worth taking seriously, not because the sleep harms you but because the need may have a cause worth investigating.

How to do it: napping inside a real schedule

Most nap advice is written for someone with their own office and a door that closes. Here is the version for an ordinary weekday.

• Decide the length beforehand, not during. Alarm at 20 minutes. One alarm, no snoozing.
• Put it between noon and three. After 3pm it costs more than it gives.
• Dark and cool beats comfortable. An eye mask in the desk drawer does more than a better chair.
• Half-reclined is fine. You do not need a bed. The car in the car park will do, provided it is parked.
• Not falling asleep is not a wasted nap. Twenty minutes with your eyes shut and no screen is still twenty minutes without input. 1177's own first suggestion for poor sleepers is precisely a relaxation exercise rather than sleep.
• Get up and go outside into the light afterwards. Daylight helps the body set its circadian rhythm, which 1177 lists among its general sleep recommendations.
• Do not nap out of habit. Nap on the days it is needed. Habit is what turns it into a daily cost instead of a tool.

Got fifteen minutes rather than twenty, with the choice between a short walk outdoors and a nap? If you sleep well at night it matters little which you pick. If you sleep badly, the walk wins, for all the reasons above.

When to seek medical advice

A nap is a tool for tiredness, not an answer to pathological sleepiness. Contact your doctor if any of the following applies:

• You are so tired during the day that you struggle to stay awake in meetings, in front of the television or in class, despite sleeping enough hours.
• You fall asleep without deciding to, particularly if it happens behind the wheel. Do not drive when you are too tired.
• You snore heavily, and someone has heard you stop breathing in your sleep. That can be sleep apnoea, described in more detail in the guide to waking at 3am.
• According to 1177, you may need help if you sleep too little over a longer period and function worse in daily life. Examples they list: it takes at least 45 minutes to fall asleep, you wake one or more times and struggle to get back to sleep, or you wake too early and cannot get back to sleep.
• The tiredness arrived suddenly or comes with other symptoms such as breathlessness, chest pain, low mood or unexplained weight change.

This text provides general health information. It does not replace assessment by a clinician, and it does not diagnose anything.

Frequently asked questions

How long should a nap be?

Sweden's public health service 1177 sets a ceiling of 15 minutes so the nap does not disturb the coming night. The laboratory lands in roughly the same place. In Brooks and Lack's 2006 experiment, 10 minutes of sleep produced improvements immediately on waking across every measure, while 30 minutes first produced a spell of reduced alertness before any benefit arrived. Allow for the time it takes to drop off, so set the alarm for 20 minutes if you need a few minutes to fall asleep. You are aiming for a short sleep, not an interrupted deep one.

What time of day should I nap?

Before 3pm, according to 1177. The reason is sleep pressure, the tiredness that builds through every waking hour and is partly discharged when you sleep. Nap in the early afternoon and the pressure has time to rebuild before bedtime. Nap at 6pm and it does not, which makes falling asleep that night harder. The doze on the sofa in front of the television at nine is the most expensive nap most people take, for exactly the same reason, even though nobody calls it a nap.

Why do I feel worse after a nap?

That is sleep inertia: a well-documented, temporary drop in cognitive performance right after waking that fades the longer you have been up. According to the 2019 review by Hilditch and McHill, it is made worse by existing sleep loss and by the time of day at which you wake. The usual reason a nap feels like a failure is that it ran too long, so you were woken out of deeper sleep rather than out of the light opening stage.

Is napping bad for you?

Observational studies show associations, not causes. A 2020 dose-response meta-analysis of 313,651 people across 20 cohorts found napping overall linked to higher all-cause mortality, hazard ratio 1.19. For incident cardiovascular disease there was no association in the pooled analysis, but there was a J-shaped curve: risk fell up to roughly 25 minutes a day and then rose sharply. The catch is that illness and fatigue are themselves reasons to sleep in the daytime. In the Swiss cohort study Häusler 2019, the raised risk among near-daily nappers vanished entirely once the analysis adjusted for major cardiovascular risk factors, plus excessive daytime sleepiness or obstructive sleep apnoea.

Should I nap after a bad night's sleep?

After one bad night, a short early-afternoon nap is reasonable. If you have persistent trouble sleeping, the answer changes: 1177 states plainly that people who struggle to sleep at night should avoid sleeping during the day, because it makes the coming night worse. In step 1 of its sleep programme this is one of the actual interventions. If you need energy in the middle of the day, 1177 recommends a relaxation exercise or a walk instead.

Does a coffee nap work?

The idea is that the caffeine starts working just as you wake. The evidence is thinner than the internet's confidence suggests. De Valck's 2003 driving-simulator study of 12 participants tested a 30-minute nap and 300 mg of slow-release caffeine separately rather than combined. Both countered sleepiness, and the caffeine effect lasted longer, which partly reflects the slow-release preparation rather than a cup of coffee. There is nothing here arguing against trying it in the daytime, but caffeine late in the day is a separate matter, covered in the guide to sleeping better after 40.

Sources for this page: 1177 on sleep difficulties for the guidance on nap length and timing and for the signs that you should seek help, and 1177's sleep programme for avoiding daytime sleep during insomnia; Brooks and Lack 2006 in Sleep (PMID 16796222) for the nap-duration comparison; Hilditch and McHill 2019 in Nature and Science of Sleep (PMID 31692489) for sleep inertia; Borbély 1982 in Human Neurobiology (PMID 7185792) and Borbély and colleagues 2016 in the Journal of Sleep Research (PMID 26762182) for the two-process model; Monk 2005 in Clinics in Sports Medicine (PMID 15892914) for the post-lunch dip; Pan and colleagues 2020 in Sleep Medicine (PMID 32858276) and Häusler and colleagues 2019 in Heart (PMID 31501230) for the epidemiology; De Valck and colleagues 2003 in Perceptual and Motor Skills (PMID 12705512) for the caffeine question.