How to Nap Effectively: Length, Timing, Caffeine Naps, and Sleep Inertia

A short nap is one of the few tools that reliably reduces sleepiness within the hour. The details decide whether you wake up sharper or groggier: how long you sleep, when, and whether you add caffeine. This page summarises the research. It is not medical advice, and frequent unexplained daytime sleepiness is worth raising with a clinician.

How long to nap

The most cited nap-length experiment is Brooks and Lack (2006). Twenty-four young adults slept about 5 hours at night, then took an afternoon nap of exactly 5, 10, 20 or 30 minutes of sleep, or no nap.

The authors judged that “the 10-minute nap was overall the most effective” of the lengths tested. Note that these times are minutes of actual sleep, so allow a few extra minutes in bed to fall asleep. The study did not test naps longer than 30 minutes.

A 2021 meta-analysis of 11 studies with 381 participants found that naps improved overall cognitive performance, especially alertness, with benefits mostly lasting up to 120 minutes. Nap length did not change the size of the effect in that analysis, and nearly all the studies were run in the lab.

When to nap

The same meta-analysis found a slightly larger effect for naps taken before 1 p.m., but the confidence interval crossed zero, so treat that as a hint rather than a rule. A practical reason to nap earlier is that a late or long nap reduces sleep pressure before bedtime. For how sleep pressure builds through the day, see the complete guide to better sleep.

Avoiding grogginess (sleep inertia)

Sleep inertia is the impaired performance and grogginess after waking. A review of countermeasures found no convincing reactive fix once you are awake. Caffeine was the best option but works best when taken before sleep. Light and cooler temperature improved how alert people felt, with less evidence on performance. The authors advise avoiding safety-critical tasks immediately after waking.

In practice that means keeping naps short enough to avoid deep sleep, giving yourself a few minutes before driving or demanding work, and planning caffeine ahead if you need it. For what happens in each sleep stage, see sleep stages and architecture.

The caffeine nap

In Reyner and Horne’s driving-simulator study, 12 sleepy drivers had a 30-minute break before a 2-hour monotonous afternoon drive. Caffeine (200 mg) followed by a nap of under 15 minutes reduced driving incidents to 9% of placebo levels, compared with 34% for caffeine alone. Even naps where people only dozed still helped. The sample was small and the drive was simulated, but the logic is sound: caffeine takes time to act, so it arrives as you wake.

Keep the caffeine dose and time of day compatible with night sleep; the caffeine cutoff guide covers timing. If you cannot fall asleep, quiet rest can still help, and the NSDR protocol describes one structured approach.

Naps on night shifts

Night workers face the hardest version of the nap question: they are short of sleep and working against their body clock. A systematic review by Ruggiero and Redeker (2014) found 13 experimental and quasi-experimental studies of naps taken during a real or simulated night shift. Most found that, despite short periods of sleep inertia right after waking, a planned night-shift nap reduced sleepiness and improved sleep-related performance. The studies were mostly small, used mixed designs, and none measured safety outcomes at work, so the authors called for larger trials before wider use. If you nap on shift, allow time for grogginess to pass before safety-critical tasks. The shift-work light protocol covers the other main lever, light timing.

When a longer nap makes sense

The short-nap advice is about alertness. For learning, longer naps that reach deeper sleep may do more. In a 2003 study in Nature Neuroscience, Mednick, Nakayama and Stickgold found that 60- to 90-minute naps containing both slow-wave and REM sleep improved learning on a visual texture task about as much as a full 8-hour night. The gain added to later night-time improvement, so performance over 24 hours matched what is normally seen over twice that time. This was one perceptual task in a lab, and a nap that long carries more sleep inertia and more risk of affecting night sleep. It may suit a day of heavy learning more than an everyday habit.

Long habitual naps and health

Cohort studies tell a different story from the lab. A 2024 meta-analysis of 44 cohort studies with about 1.86 million participants linked habitual napping with higher all-cause mortality and cardiometabolic risk, and lower risk of cognitive impairment and sarcopenia. The split was by length: naps of 30 minutes or longer carried higher risk, while shorter naps showed no significant risk. An earlier meta-analysis of 11 cohorts found daily naps of 60 minutes or more associated with 27% higher all-cause mortality.

These studies cannot show that long naps cause harm. A review of the nap paradox notes that the associations persist after statistical adjustment, but long naps may also reflect poor night sleep, a sleep disorder or existing illness. A Mendelian randomization study in the UK Biobank, which uses genetic variants to reduce confounding, found habitual napping associated with modestly larger total brain volume but no difference in reaction time or visual memory.

A simple nap routine

• Keep it short. Aim for 10 to 20 minutes of sleep, which matches the Brooks and Lack results and stays under the 30-minute mark linked with higher risk in cohort studies.
• Go early. Early to mid-afternoon fits the natural post-lunch dip and protects night sleep.
• Set an alarm. It keeps you from sliding into deep sleep.
• Add caffeine only if it fits. Take it just before lying down, and only early enough in the day.
• Give yourself a buffer. Allow time to shake off grogginess before driving or high-stakes work.

Limits of the evidence

The nap-length data come mostly from small lab studies of healthy young adults after restricted sleep. They may not apply to well-rested people, shift workers or older adults. The health data are observational and rely on self-reported napping.