lifescore Edge7 min read

Can a short nap improve performance?

A 381-person meta-analysis found a small post-nap cognitive benefit, strongest for alertness; nearly all tests were in labs, the average nap lasted 55 minutes, and immediate results were inconsistent.

Read the evidence

August 4, 2026

A shaded rest chair is separated by a hazed band from a clear sunlit work surface in a quiet afternoon studio.

The power-nap promise sounds unusually precise: close your eyes for a few minutes, wake up sharper and return to work ahead. The evidence is less tidy.

A daytime nap can improve later cognitive performance on average. In the most relevant meta-analysis, the pooled effect was small and clearest for alertness. But almost every experiment happened in a laboratory, the average nap was much longer than most readers would call short, and the first minutes after waking were the least reliable part of the record.

That makes the useful answer conditional. A nap can create a later performance window. It does not guarantee an immediate one.

The best pooled result is positive, small and mostly about alertness

In 2021, researchers combined 11 controlled studies with 381 working-age participants. The studies compared cognitive testing after a daytime nap with testing after a no-nap control condition.

The pooled post-nap difference was a standardized mean difference of 0.18, with a 95% confidence interval from 0.09 to 0.27. Standardization lets different tests share one scale. An estimate of 0.18 is a small average shift, not a transformation in how every participant performed.

The cognitive categories did not move equally:

Outcome	Standardized mean difference	95% confidence interval	Reading
Alertness	`0.29`	`0.10` to `0.48`	Clearest pooled signal
Executive function	`0.23`	`0.00` to `0.47`	Borderline lower bound
Memory	`0.11`	`-0.01` to `0.24`	Interval includes no average difference
Overall cognition	`0.18`	`0.09` to `0.27`	Small pooled advantage

This distinction matters. Feeling and responding more alert is not identical to remembering more, reasoning better or producing more valuable work. A headline that says performance improved can conceal which performance measure actually carried the result.

The review called the naps short; the average was 55 minutes

The paper’s title refers to a short daytime nap. Its included sleep durations ranged from 15 to 90 minutes and averaged 55.4 minutes.

That is not a semantic footnote. It means the pooled estimate cannot tell a reader that a 10-, 20- or 26-minute nap will reproduce the average effect. The review’s meta-regression did not detect a duration effect, but comparisons across a small set of different studies are not the same as randomizing people to several nap lengths under one protocol.

The popular duration question therefore remains open. Search results often offer one memorable number. The primary synthesis offers a range, an average and a warning about immediate testing.

The first 30 minutes after waking are the uncertain zone

Sleep inertia is the temporary reduction in alertness or performance after waking. It can feel like grogginess, but its practical cost is not limited to a feeling: reaction speed, judgment or task accuracy can briefly lag as waking systems come back online.

In the meta-analysis, the pooled result for tests within 30 minutes of waking was positive in the main model. It did not remain significant after outlying studies were removed. By contrast, benefits were more consistent in later testing windows, particularly between roughly 30 and 120 minutes.

A separate review focused specifically on naps of 30 minutes or less found mixed evidence about slow-wave sleep and sleep inertia. Prior sleep, clock time and the sleep reached during the nap all changed the likely response.

So under 30 minutes is not a force field against grogginess. It is one input among several.

A 2025 experiment separated feeling better from performing better

A newer randomized study tried to wake people at a more biologically precise moment: nine minutes after detecting N2 sleep.

Eighty-one healthy adults were assigned in equal groups of 27. One group used an automatic system based on an ear-worn blood-flow sensor. A second was awakened manually using real-time polysomnography. A third rested without sleeping. Participants then repeated digit-symbol, visual-detection, sleepiness and fatigue measures across six post-nap sessions.

The automatic system missed its intended timing often. Its reported first- generation N2-detection accuracy was 25.9%. Offline analysis showed that the alarm sounded an average of 19.6 minutes after N2, not nine. Total sleep averaged 22.6 minutes, and 63% of that group reached N3, the deeper stage the protocol was trying to avoid.

Participants in the automatic arm reported nominally lower sleepiness and fatigue than the rest group. The study was not blinded, and these subjective analyses were not protected by the multiplicity control applied to the primary analysis. The primary objective result did not improve: digit-symbol performance was not significantly different from rest in any post-nap session. Visual-detection misses and reaction time also did not differ.

The manually awakened group slept for an average of 11.8 minutes and rarely reached N3. Its nominal, unadjusted digit-symbol advantage appeared only in the sixth post-nap session—roughly 95 to 155 minutes after the nap—not across the earlier sessions. Its visual-detection outcomes did not improve.

The researchers later retrained the N2-detection system and tested its timing accuracy in a separate 50-person validation. Detection improved under the paper’s wider timing tolerances. That second experiment did not wake people for a nap-versus-rest comparison and did not retest cognitive performance, so it cannot repair the first experiment’s objective null result.

The experiment does not erase the meta-analysis. It clarifies what a pooled headline can hide. A nap may nominally reduce subjective sleepiness without moving the objective task being tested. A task difference may appear late rather than immediately. And a system designed to optimize sleep stage can fail if it cannot detect that stage accurately.

KYOCERA supported the study, and a KYOCERA employee helped design it and developed the automatic-awakening system. Public and academic grants also supported the work. That provenance does not determine the findings; it does make the device-specific null and accuracy results especially important to keep visible.

Athletic performance is a different evidence lane

The word performance also reaches sport, where the outcome might be sprint power, endurance, reaction time, soreness or perceived effort.

A 2021 athletic review called the evidence equivocal. Many studies reported benefits after normal sleep or sleep loss, but others found no benefit or a temporary deterioration. The pattern depended on prior night sleep, nap length and the recovery interval before testing. Sleep inertia was one plausible reason a nap could help later yet hurt an early measurement.

That record does not justify importing an athletic result into office work. A sleep-deprived athlete preparing for a timed physical test and a rested worker preparing for a negotiation are not the same population, outcome or cost of an early error.

What the evidence changes

The nap question is better framed as a timing system than as a universal duration hack.

Decision point	What the evidence supports	What remains unproved
Outcome	Alertness has the clearest pooled cognitive signal	Every form of memory, reasoning or productivity improves
Duration	Studies span many durations	One universally optimal number
Immediately after waking	Results are unstable and sleep inertia can occur	Instant readiness for high-stakes work
Later after waking	Small average cognitive benefits can persist	The same effect in every real workplace
Subjective recovery	Sleepiness and fatigue can improve	Objective task performance necessarily follows
Sport	Some contexts show recovery or performance gains	Transfer to rested non-athletes or knowledge work

The practical implication is not a dosing-style nap prescription. It is a design constraint for any situation where errors matter: do not assume the moment after waking is the benefit window simply because a later average is positive.

Primary sources

Independent editorial summary. The authors are not affiliated with LifeScore.

The lifescore take

A nap is not stored performance waiting to be unlocked by one magic alarm. It is a transition through sleep stage, waking inertia and recovery into a possible later advantage. The strongest current answer is modest. Daytime napping can improve later cognitive performance on average, particularly alertness. The evidence is too laboratory-heavy and too heterogeneous to promise an immediate productivity boost or one ideal short duration. The sharper future study would happen in a real work setting, prespecify one meaningful task, compare defined nap and recovery intervals, and measure both subjective readiness and objective errors. Until then, the minutes after waking belong inside the experiment, not outside the claim.

Article link