How to Sleep After a Night Shift Sound, light and timing, in the order you meet them.

Sleep after a night shift is shorter because you are trying to sleep when your body clock is pushing hardest for you to be awake. Sleep is governed by two things: how long you have been awake, which builds pressure to sleep, and a circadian signal on a roughly 24-hour cycle that runs whether you worked or not. After a night shift the first is high and the second is against you, so you fall asleep readily and wake at eleven or noon with your body insisting the day has started.

The clock is also stubborn. A 2022 review by Boivin, Boudreau and Kosmadopoulos in the Journal of Biological Rhythms (PMID 34969316) covers both simulated night-shift experiments and field studies of real shift workers and finds the same thing in both: the central clock, measured by melatonin and cortisol rhythms, shows no substantial shift over multiple days of night work. Most people who work nights do not adapt to them; they live, physiologically, on a day schedule and sleep in its gaps.

So the things that help are the things that act on the two levers — light, the clock’s main input, and timing, which decides how much sleep pressure you bring to bed — plus darkness and noise, which stop the daytime world from cutting a short sleep shorter still. Audio is the last of those, and it is the one this site makes, so we will be plain about what it does and does not do.

If your problem is a time-zone crossing rather than a roster, the clock science is the same but the tactics differ; jet lag recovery covers that case and this page does not repeat it.

Two decisions made at work decide a lot of what happens in bed. The first is when you stop drinking coffee. The second is what light you are under, and when.

Caffeine: stop earlier than feels necessary

The best-known timing study is Drake and colleagues (2013) in the Journal of Clinical Sleep Medicine (PMID 24235903), which gave a fixed 400 mg dose of caffeine — roughly two large coffees — at bedtime, three hours before bed, or six hours before bed, against placebo, and measured sleep at home with a portable monitor. All three timings disrupted sleep measurably; the six-hour dose still cut total sleep time enough that the authors’ conclusion was to refrain from substantial caffeine for at least six hours before sleep.

Translate that to a shift ending at 07:00 with sleep at 08:00: the last coffee is around 02:00, earlier than most night workers stop. The dose in that study was large and smaller amounts disrupt less, but the direction is clear. Caffeine on the front half of the shift, water and food on the back half.

What caffeine is good for is the shift itself. A Cochrane review by Ker and colleagues (2010) (PMID 20464765) found thirteen trials of caffeine in shift workers and people with jet lag; none measured injuries, but the two that measured errors found caffeine reduced them against placebo. So the advice is not “avoid caffeine” — it is front-load it.

Light: bright in the first half, dimmer towards the end

Research on shift workers has tried to use light two ways: bright light during the night to push the clock later, and darkness after the shift to stop it being pulled back. The Chicago simulated-night-shift series by Smith, Eastman and colleagues is the clearest worked example. In the 2008 study in Physiology & Behavior (PMID 18675836), experimental groups received intermittent bright-light pulses during the night shift, wore dark sunglasses outside, slept in dark bedrooms at fixed times and got outdoor light on waking; their melatonin onset moved several hours later than the control group’s, and they did better on a reaction-time task at night. Study 4, in the Journal of Biological Rhythms in 2009 (PMID 19346453), used four 15-minute light pulses per shift with the same routine and reached its target clock position.

Two caveats. Those were simulated shifts with healthy young volunteers, lit by light boxes rather than a ward or a warehouse. And the 2016 Cochrane review of non-drug interventions for shift workers, by Slanger and colleagues (PMID 27549931), pooled what trials there were — ten on bright light — and graded almost all of it low or very low quality. Bright light at night reduced on-shift sleepiness by under one point on a nine-point scale in two pooled studies; the one trial of bright light plus sunglasses at dawn found no significant difference; and the review concludes it cannot yet say whether these interventions improve shift workers’ sleep. That is why the advice here is phrased as cheap things that follow from how the clock works, not as proven treatments.

What follows from the mechanism: keep the workplace lighting bright early in the shift, and avoid the brightest light in the last hour or two. A phone at full brightness at 06:30 is a small dose of exactly the wrong signal.

The walk from the building to the car is the most important light exposure of your day, and it is the one nobody plans for. Morning daylight is far brighter than any indoor lighting, it arrives exactly when your clock is most sensitive to being pulled towards a day schedule, and it tells your body the day has begun just as you are trying to end it.

The intervention is embarrassingly simple and is the one the Eastman studies above built into every experimental group: dark sunglasses from the door to the bed. Wrap-around if you can get them, because light reaches the eye from the sides. Put them on before you step outside, not in the car park. If you cycle or walk, the same applies, with the obvious caveat that you still need to see the road.

Then there is the drive. Drowsy driving after a night shift is the part of this that can kill you: if you have been awake since the previous afternoon and are driving at the hour your clock is lowest, you are impaired in the way that matters, whatever you feel like. If you find yourself drifting, stop and nap somewhere safe; if your employer has a rest room, a twenty-minute nap before you leave is cheaper than any coffee. The nap timer on this site is built for exactly that length.

A daytime bedroom has three problems a night-time bedroom does not: it is bright, it is warm, and the world outside it is awake. Fix them in that order.

Dark

Blackout blinds or curtains, fitted so they overlap the frame; the gap at the edge of a blind is where most of the light comes in. Failing that, a sleep mask that seals at the nose does the same job for the eyes, which is what matters for the clock. Cover anything with a standby light. Darkness is the other half of the sunglasses: it keeps the morning from resetting your clock while you sleep.

Cool

Body temperature falls into sleep, and a daytime room warms up from mid-morning. Cool the room before you get in rather than after, and sleep under less than you would at night. A fan is doubly useful, because it is also a continuous noise source — which is the third problem.

Masked: what brown noise is honestly for

Daytime loudness has a shape. Traffic, a bus pulling away, a washing machine next door, footsteps overhead: most of what comes through walls and windows is low-frequency, because low frequencies pass through building materials far more easily than high ones. From inside you hear the rumble, not the hiss.

Masking works on overlap. A steady sound hides an intermittent one only where the two share frequencies and the steady one is at a comparable level at your ear. Brown noise puts most of its energy at the bottom of the spectrum, rolling off at 6 dB per octave, which is why it is the right default here: it sits in the same band as the rumble and raises the floor under it, so a passing car becomes a small change in a sound you have stopped noticing instead of a silence being broken.

Be precise about what masking does not do. It does not delete a sound; it reduces the contrast between the sound and the background, and the contrast is what wakes you. A slammed door, a doorbell or a shouting child is a loud transient and will come through any noise you could sleep under — turning the noise up until it covers those is how people end up sleeping under something louder than the problem. And brown noise rolls off exactly where voices live, between a few hundred and a few thousand hertz, so if the intrusion is a television through the wall or a household having its day, pink noise, which keeps more energy up there, usually does the job better. That is why the player above offers both and nothing else.

Three settings that matter. Volume low — raise it from quiet until the street is blunted, and stop there; if the noise is the loudest thing in the room it is too loud for seven hours. Speaker, not earbuds — noise needs no stereo separation, and a small speaker between you and the window does more at the same volume than anything in your ears. Continuous — the one place the free web tool is the wrong size. It stops at twenty minutes, which covers falling asleep and nothing after; for the whole day sleep you need a fan, a noise machine, or the app, which runs long sessions with the screen off.

On the research: the evidence that continuous noise improves sleep is modest, from small trials elsewhere, and untested in shift workers. The acoustic claim — a steady low-frequency bed makes intermittent low-frequency noise less likely to wake you — is physics, not a clinical promise, and it is the only claim this page makes. If you want a beat underneath, binaural beats while sleeping explains why we would not run one all day, and sleep frequencies gives the delta settings for the first twenty minutes.

The single most protective thing you can do is to sleep at the same clock time on every day you can, including some of the days off. This is the idea of anchor sleep, and it is older and better supported than most shift-work advice.

Minors and Waterhouse studied it in an isolation unit and published the constant-routine evidence in the Journal of Physiology in 1983 (PMID 6663508): body-temperature and renal rhythms stayed “anchored” to a 24-hour period if four hours of sleep were taken at the same time every day, even when another four hours were taken irregularly. You do not need your whole sleep to be regular for your clock to stay stable; you need a block of it to be.

That gives three workable patterns; which suits you depends on the roster and the house you go home to.

1. Straight away Home, sunglasses off at the door, bed within the hour. This spends your sleep pressure while it is highest and is the right default for a run of consecutive nights. Expect to wake after five or six hours regardless; that is the clock, not a failure.
2. Split A main sleep in the morning, then a second, shorter sleep in the late afternoon before the next shift. The pre-shift nap is the part people skip, and the part that keeps the last hours of the night from falling apart; the next section covers it.
3. Anchored Pick a block of three or four hours — say 09:00 to 13:00 — and protect it on working days and days off, when you might sleep 03:00 to 11:00 rather than a conventional night. The rest floats around the roster. This is what the Eastman studies were aiming at: a compromise clock position, neither fully night- nor day-adapted, so that both the shift and a day off are survivable.

Days off are where this is won or lost. The temptation is to flip completely — sleep at night, be up all day — then flip back for the next run of nights, so every run starts already misaligned. The compromise in Smith, Fogg and Eastman’s 2009 study (PMID 19346453) exists because, in their words, few night workers would want full adaptation: it leaves them out of step with everyone on their days off. A late night and a late morning on days off, rather than a conventional one, kept their subjects nearer the target. Rotating shifts are harder than permanent nights for exactly this reason, and on a fast rotation there may be no stable pattern to find; then the anchor block is still the thing to protect.

Two more rules from the mechanism. Do not set an alarm for the day sleep unless you must; it is short enough without cutting it. And when you wake, get light — open the blinds, go outside — because light on waking is the “light brake” that stops the clock drifting so late that days off become unliveable.

A nap before a night shift is not a luxury; it is the second half of a split sleep, and it is the part of this page with the most direct evidence. The Cochrane review of drug interventions for shift work by Liira and colleagues (2014) (PMID 25113164) included one trial of caffeine plus a pre-shift nap, which reduced sleepiness during the night on the Karolinska scale — low-quality evidence, from a single trial, but pointing the right way.

On naps during the shift, Ruggiero and Redeker’s 2014 systematic review in Biological Research for Nursing (PMID 23411360) found thirteen mostly small studies, and most reported that a planned nap on the night shift reduced sleepiness and improved performance, despite a short period of grogginess afterwards. None measured workplace safety, and the authors called for larger trials before wide adoption. The Slanger Cochrane review above was more cautious: its pooled estimate for a single on-shift nap, by reaction time, did not reach significance. Read together, the honest summary is that a short nap probably helps you feel and perform better for the rest of the night, and the size of the effect is not well measured.

The grogginess is the thing to manage. Keep the nap short enough to stay out of deep sleep — twenty minutes is the usual compromise — and allow a few minutes of light and movement before doing anything that needs you sharp. The power nap page goes through the length question properly and has a timer that fades out at twenty minutes, which is why the web cap is a feature there and a limitation here.

Timing the pre-shift nap: late afternoon or early evening, finishing at least an hour before you leave. Too close to the main day sleep and it steals from it; too close to the shift and you clock in groggy. Roughly 17:00 to 19:00 for a 23:00 start.

People reach for a pill when the sleep is short, and the trials say the pill is the smallest lever on this page. The Liira Cochrane review (PMID 25113164) added it up across fifteen placebo-controlled trials and 718 participants.

• Melatonin after the night shift may lengthen daytime sleep by about 24 minutes against placebo, from seven trials graded low quality, with no dose-response effect and no change in how long it took to fall asleep. Twenty-four minutes is not nothing on a six-hour sleep, but it is less than a blackout blind.
• Hypnotics — zopiclone, in the one usable trial — did not lengthen daytime sleep against placebo.
• Modafinil and armodafinil, taken before the shift, reduced sleepiness by about one point on a ten-point scale in people with a diagnosed shift-work sleep problem, with moderate-quality evidence, at the cost of headache, nausea and raised blood pressure in the trials and severe skin reactions reported after marketing. There were no trials in shift workers without the diagnosis.

So: melatonin is a small, probably real effect on length; sleeping tablets are not supported for this; and the alertness drugs are prescription treatments for a diagnosed disorder, not a tool for a bad week. Any of them is a conversation with a doctor, not a decision made from this page, and that conversation is the subject of the next section.

Short daytime sleep is normal for night work. Being unable to sleep at all, or to stay awake on the shift, is not, and it has a name. Shift work disorder is a recognised diagnosis: insomnia or excessive sleepiness that tracks the work schedule and goes beyond what the schedule itself explains. Drake and colleagues described it in Sleep in 2004 (PMID 15683134) from a community sample of 2,570 people in Detroit: night and rotating workers who met the criteria had higher rates of ulcers, sleepiness-related accidents, absenteeism, depression and missed family life than shift workers who did not, and in most cases worse outcomes than day workers with the same symptoms. It is a condition with a treatment pathway, not a character flaw.

See a doctor if any of the following is true for more than a few weeks:

• You cannot get more than a few hours of sleep in the day despite a dark, quiet room and a protected sleep window.
• You are falling asleep, or fighting not to, during the shift or on the drive home — the second of those is a reason to go this week.
• Your mood has gone: flat, irritable, tearful, or anxious in a way that does not lift on days off. Sleep loss and low mood feed each other; anxiety at night covers that loop.
• You snore heavily, wake gasping, or are exhausted even after a long sleep. Sleep apnoea is common, under-diagnosed and treatable, and no shift management compensates for it.
• You are using alcohol to get to sleep. It works for the first hour and wrecks the rest.

Go with the facts: your roster for the last month, when you actually slept, how long, and what you have tried. A two-week sleep diary is worth more to a clinician than any description. Nothing on this page, and no sound anywhere, is a substitute for that appointment.