🌙 Quitude

Learn › CBT-I

Why do I wake up at 3am and can't get back to sleep?

Waking in the small hours is not, by itself, a malfunction: human sleep gets lighter in the second half of the night, and everybody surfaces briefly between sleep cycles. What turns a normal 3am arousal into an hour of staring at the ceiling is usually what happens next — the waking gets registered as a problem, arousal climbs, and the bed slowly becomes a place associated with being awake and frustrated. For many people the way out is not more effort to fall asleep; it is closer to the opposite, and it has a well-tested name: stimulus control.

The short version: you are supposed to wake up

Sleep is not one solid block. It moves through cycles — roughly ninety minutes each, though this varies between people and across the night — and at the end of every cycle you come close to the surface. Most of those brief arousals are never remembered. You shift position, pull the duvet across, and you are gone again. They show up on every normal sleep recording, in every healthy sleeper.

So the first honest answer to "why do I wake up at 3am" is a slightly boring one: you have woken in the night on every night of your life. On the good nights, you simply did not notice. What is different on a bad night is not the number of awakenings — it is what one of them gets turned into.

What your body is doing in the small hours

Deep, slow-wave sleep is front-loaded. Your body takes the sleep it most needs early, which is why the first few hours usually feel solid and the second half feels flimsier. By around 3am, most of that deep sleep is behind you and the night is dominated by lighter stages and REM — stages that sit much closer to consciousness and are far easier to wake out of. It is also why a 3am waking so often arrives with a vivid dream still hanging in the room.

At the same time, the sleep pressure you built up across the day has largely been discharged. Your core body temperature is near its low point and about to begin rising, and the body's cortisol rhythm starts its climb toward morning in the early hours. None of this is a fault. It is a healthy system doing exactly what it does as a night winds toward its end.

Plenty of ordinary things can turn one of those invisible arousals into a fully conscious one:

Why you can't get back to sleep — the part that's actually the problem

The waking is normal. The stretch that follows is the insomnia. The moment you register that you are awake, an appraisal happens — this again — and the appraisal is a threat signal. Heart rate nudges up, the mind switches from idle to on, you check the clock, you do the arithmetic, you forecast tomorrow's meeting through the lens of a bad night. That cascade has a name in the research literature: hyperarousal. And arousal is the precise opposite of the state sleep requires. The alarm system has switched itself on, and it does not know the difference between 3am and a real threat.

There is a second, slower process running underneath. If you regularly spend long stretches awake, tense and frustrated in bed, your brain learns the association. Bed stops being a reliable cue for sleepiness and starts being a cue for wakefulness — the same way a desk can become a cue for concentration. This is conditioned arousal, and undoing it is the entire rationale behind stimulus control therapy, one of the oldest and best-supported components of CBT-I (cognitive behavioural therapy for insomnia).

It is worth being precise about how the evidence sits, because it shapes what this page can honestly promise. The American Academy of Sleep Medicine's 2021 guideline on behavioural treatments (Edinger et al., Journal of Clinical Sleep Medicine) gives multicomponent CBT-I the only strong recommendation in the document. Stimulus control on its own also gets a recommendation, a more tentative, conditional one — it is the single most useful thing to do at 3am, and it works best as part of the full programme rather than as a standalone trick.

The third piece is the effort paradox. Sleep cannot be produced by trying. Every other problem in your life responds to effort; this one tends to get worse under it. Which is why the instructions below can feel counterintuitive — they ask you to stop attempting to sleep.

Put the clock away

Clock-watching converts a wake-up into a performance metric, and the number is never reassuring. Whatever it says, the response is the same: a fresh calculation of how much sleep is left and how bad tomorrow will be. That calculation is pure arousal fuel.

Turn the clock to face the wall. Put the phone face down, out of reach, ideally in another room. Keep the alarm — you are not risking oversleeping, you are just declining to check the score. As a bonus, you stop generating the detailed ledger of awakenings that feeds the next day's catastrophising.

Before you change anything: who should talk to a doctor first

Everything in the next two sections works partly by raising sleep pressure — the biological drive to sleep that builds the longer you stay awake. That is the mechanism, and it means the first week or two often brings more daytime sleepiness, not less. For many people that is a manageable trade for a few weeks. For some people it is not, and it is better to know which group you are in tonight rather than afterwards.

Talk to a doctor before changing your sleep schedule, restricting your time in bed, or starting the get-out-of-bed steps if any of these apply to you:

The safety rules that outrank the protocol

One driving line, and it is for every reader, not only people who drive for a living: while you are sleepier than usual, do not drive or operate machinery when you feel sleepy. Drowsiness at the wheel is a genuine hazard, and "I am nearly home" is not a plan. If you are fighting to stay awake in the car, stop the car.

There are also points at which the right response is to stop and get advice rather than push on. Widen your sleep window — give yourself more time in bed — and speak to a clinician if any of the following happen:

  • You are falling asleep unintentionally during the day, or having trouble staying awake when you need to be alert
  • You nod off, or have to fight to stay awake, while driving, at work, or while caring for someone
  • Your mood becomes elevated, agitated, wired or unusually fast — a particular concern if you have any history of mania or hypomania
  • Your mood drops significantly, or hopelessness sets in
  • Your sleep or your daytime functioning gets meaningfully worse and stays worse

What to do at 3am tonight

This is stimulus control, in practice. It is simple to describe and genuinely hard to do for the first few nights, and when it works, it works by rebuilding the association between your bed and sleep.

Two things before you try it. First, if you get out of bed in the night, put a light on and move carefully — these instructions raise fall risk if you are older, unsteady, dizzy on standing, or taking sedating medication. Second, if you have a seizure disorder, bipolar disorder, a parasomnia, untreated or suspected sleep apnoea, narcolepsy or existing daytime sleepiness, a safety-critical job, or you are pregnant, talk to a doctor before changing your sleep schedule. The full list is in the section above, and it applies to these steps too, not only to formal sleep restriction.

If it happens most nights, look at your time in bed

A note on who this part is for. Restricting time in bed is a treatment for chronic insomnia — conventionally, trouble sleeping at least three nights a week for three months or more, with knock-on effects during the day. If you have had a rough fortnight after a stressful event, a schedule change, a new baby or a trip across time zones, you probably do not need it. The stimulus control steps above are usually enough, and short-lived bad patches often settle by themselves.

For long-running insomnia, the instinctive response to broken sleep is to spend longer in bed — go up early, lie in at the weekend, give sleep more opportunity to happen. It is completely understandable and it reliably backfires. Spreading the same quantity of sleep across a longer window makes it thinner and more broken, and manufactures more of exactly the wakeful, frustrated bed-time that drives conditioned arousal.

CBT-I does the opposite. Using one to two weeks of sleep diary data, it temporarily matches your time in bed to the sleep you are actually getting — with a floor. Standard protocols never prescribe less than roughly 5 to 5.5 hours in bed, however little sleep the diary shows, because below that the daytime sleepiness stops being uncomfortable and starts being dangerous. That floor exists for safety, so do not out-strict it: if the arithmetic says four hours, you use the floor instead. Quitude's engine enforces a floor of 5 hours 30 minutes and will not prescribe a shorter window under any circumstances — deliberately more conservative than the supervised clinical protocols, because nobody is reviewing you week to week.

The window is then widened in small steps as sleep efficiency — time asleep divided by time in bed — improves. Quitude reviews this once a week against seven days of diary data, following Spielman's convention: at 90% efficiency or above the window widens by 15 minutes; between 85% and 90% it holds; below 85% it tightens by 15 minutes, and never below the floor. Weekly rather than nightly matters — one bad night is noise, a week is a signal. Quitude's seven-week CBT-I programme is built around that loop, and it is the free part of the app.

One caveat that deserves respect rather than small print. Deliberately compressing time in bed causes daytime sleepiness in the first week or two, and that matters a great deal if you drive, operate machinery, or care for anyone. Work through the contraindications listed above and speak to a doctor first if any of them apply to you. Keep the stop rules in mind once you have started, too: if you are falling asleep unintentionally, nodding off at the wheel or at work, or your mood swings up or drops down, add time back to the window and speak to a clinician rather than pushing through. Feeling worse in those specific ways is not evidence the treatment is working.

CBT-I as a method has strong guideline support. An app delivering that protocol as unsupervised self-help is not the same thing, and it would be dishonest to blur them: Quitude has not been through clinical trials of its own. The trial evidence for digital CBT-I comes from clinician-delivered programmes and from separately tested products such as Sleepio and Somryst — not from Quitude. It is a faithful delivery of a well-evidenced protocol, not a validated treatment, and not a replacement for care from a clinician.

What tends not to fix a 3am waking

Sleep hygiene alone. Cool room, no late caffeine, no screens, consistent bedtime — this is sensible supporting context, and the AASM's 2021 guideline is specific that it is not adequate as a stand-alone treatment for chronic insomnia. If you have already done everything on the hygiene checklist and still wake at 3am, that is not a personal failure. You were handed an incomplete tool.

Sleeping pills, at least for the long run. The AASM's 2017 medication guideline (Sateia et al., Journal of Clinical Sleep Medicine) made fourteen recommendations and graded every one of them weak, with evidence quality mostly low or very low; melatonin sits among the recommendations against use for chronic insomnia. A 2024 network meta-analysis (Furukawa et al., Psychiatry and Clinical Neurosciences; 13 trials, 823 participants) found that among people who started on CBT-I, around 41% were in remission at long-term follow-up — median 24 weeks — compared with about 28% of those who started on medication, and fewer dropped out (roughly 21% versus 39%). Starting on both together showed no clear advantage over CBT-I alone. Those were treatment-initiating arms, so people could change course afterwards, and 13 trials is a modest evidence base for a head-to-head claim — but the direction is consistent. The honest framing: medication works while you take it, and for many people the CBT-I skills keep working after the programme ends. That 2017 guideline reserves long-term hypnotic use for people for whom CBT is inaccessible or ineffective.

Never stop or change a prescribed sleep medication on the strength of an article. Some sleeping pills need a supervised taper, and stopping abruptly can cause a withdrawal reaction that is a medical risk, not just a bad week of sleep. That conversation belongs with whoever prescribed it.

The nightcap. A drink genuinely does make it easier to fall asleep, and then fragments the second half of the night as it is metabolised. It is one of the most efficient ways to produce a 3am wake.

When to get it checked by a doctor

Behavioural techniques treat insomnia. They do not treat the other things that wake people at 3am, and some of those need a proper look.

If any of the following fit you, book an appointment rather than starting a self-help programme on your own:

One last thing, for tonight

The target is not a perfect, unbroken eight hours. Chasing that specific image is itself one of the more reliable ways to make insomnia worse, because it turns every normal arousal into evidence of failure. Humans have always slept imperfectly, and one rough night is survivable — you will probably be tired tomorrow, and tired is usually not the catastrophe that 3am predicts it will be.

One exception, and it is not negotiable: sleepiness behind the wheel or at a machine is a genuine hazard, not a catastrophic thought to be talked down. If you are fighting to stay awake, do not drive, and take a short early nap instead even if you are mid-programme and the rules say no naps.

The realistic goal of CBT-I is not that you stop waking in the night. For many people it is that when they do wake, it stops mattering: they notice, they turn over, and the machinery of alarm never starts up. When it works, that shift takes a few weeks of fairly unglamorous practice, and the gains tend to hold after the programme is over — that is what the group data shows, though no programme can promise a particular result on a particular timeline for any one person.

This article is general information, not medical advice. If your sleep problems continue, or you are worried about them, please speak to a doctor or a sleep specialist.

Frequently asked questions

Why do I wake up at exactly the same time every night?

Two things are usually going on. Your sleep cycles and circadian rhythm run on a fairly regular schedule, so the lighter, easier-to-wake portions of the night arrive at roughly the same clock time each night — and if you have been waking and lying there for a while, your brain learns that pattern too, much like waking a few minutes before an alarm. Memory plays a part as well: you do not remember the arousals you slipped straight back through, only the one that turned into an event. It is not a signal about a specific organ or a particular hour having meaning.

Does waking at 3am mean my cortisol is high or my liver is struggling?

You will see both claims a lot, and neither is a useful explanation for most people. Cortisol does begin rising in the small hours — that is the normal morning rhythm in every healthy person, not a fault. The traditional "organ clock" idea that a particular hour maps to a particular organ is not supported by sleep physiology. If you have other symptoms — unexplained weight change, night sweats, persistent pain, palpitations — see a doctor about those symptoms directly. Otherwise, sleep architecture plus arousal explains a 3am waking far better.

How long should I lie there before I get out of bed?

The usual guidance is around fifteen to twenty minutes, but judge it by feel rather than timing it, because checking the clock is itself part of the problem. The real cue is your state: if you are calm and drowsy, stay in bed. If you are wide awake, tense, or your mind has started running the arithmetic of tomorrow, get up, go somewhere else, and do something quiet and dull until you feel sleepy again. Two safety notes: put a light on and move carefully, since getting up in the dark raises fall risk if you are older, unsteady, dizzy on standing or on sedating medication — and check with a doctor before changing your sleep schedule if you have a seizure disorder, bipolar disorder, a parasomnia such as sleepwalking, untreated or suspected sleep apnoea, narcolepsy or existing daytime sleepiness, a safety-critical job, or you are pregnant.

Should I sleep in or nap the next day to catch up?

Usually not. Sleeping late or napping heavily spends the sleep pressure you need for the following night, which makes another fragmented night more likely — it is a common way a single bad night turns into a run of them. Keep your rise time fixed, get some daylight in the morning, and accept a tired day. Safety is the exception that outranks the rule: if you are too sleepy to drive, work or care for someone safely, take a short nap early in the afternoon rather than pushing through. A short early nap costs you less than a long or late-evening one, and far less than a drowsy drive.

Will melatonin or a sleeping pill stop me waking at 3am?

Possibly in the short term, but the evidence does not favour it as a long-term answer. The AASM's 2017 medication guideline graded all fourteen of its recommendations as weak, with mostly low or very low quality evidence, and placed melatonin among the recommendations against use for chronic insomnia. In a 2024 network meta-analysis (Furukawa et al.; 13 trials, 823 participants), about 41% of people who started on CBT-I were in remission at long-term follow-up — median around 24 weeks — versus roughly 28% of those who started on medication, with fewer dropouts, and starting on both offered no clear advantage over CBT-I alone. Medication works while you take it; for many people the behavioural skills keep working afterwards. If you are already on a prescribed sleep medication, do not change or stop it without speaking to your prescriber — some need a supervised taper, and stopping abruptly can cause a withdrawal reaction that is a medical risk in its own right.

How do I know when to stop and speak to someone instead of pushing on?

Some extra daytime sleepiness in the first week or two is expected when you tighten your sleep window, and it is not a sign of failure. But there are clear points where the right move is to add time back to your window and get advice rather than persist: if you are falling asleep unintentionally during the day, if you nod off or fight to stay awake while driving, at work or while caring for someone, if your mood becomes elevated, agitated or wired, or if your mood drops significantly. None of those is the treatment working. If you have thoughts of harming yourself, get help immediately — in the US call or text 988, in the UK call 111 or Samaritans on 116 123, and elsewhere your local emergency number.

Try the program free

Quitude runs the CBT-I protocol from your own sleep diary — the program is free, there is no account, and your data stays on your device.

Start the CBT-I program

Keep reading