Sleep restriction for insomnia: how to do it, step by step

Sleep restriction for insomnia means spending less time in bed for a few weeks, roughly only the hours you actually sleep, then adding time back 15 minutes at a time as your sleep fills the window. It sounds backwards, but it is one of the best-studied habit treatments for insomnia and a core part of cognitive behavioral therapy for insomnia (CBT-I).

It works by building up sleep pressure and teaching your brain that bed is for sleeping, not lying awake. The first one or two weeks are usually harder, because you will feel sleepier during the day. Below: why it works, what the trials found, who should only do it with a clinician, the exact steps with clock times, a table of the weekly rules, and what to do when it stalls.

What is sleep restriction therapy?

Sleep restriction therapy was described by the psychologist Arthur Spielman and colleagues in 1987. The idea came from a simple observation: people with insomnia often spend far longer in bed than they sleep. Someone who sleeps 5 hours and 30 minutes might spend 8 or 9 hours in bed, hoping more time will give them more sleep. What it gives them is 3 hours of lying awake, spread across the night.

The therapy shrinks the time in bed, which clinicians call the sleep window, to match the sleep you are really getting. You keep one wake-up time every day of the week. You go to bed later, not get up earlier. Once you sleep through most of the window, you widen it.

It is rarely used alone today. In CBT-I it sits next to stimulus control (leave the bed when you can't sleep), a wind-down before bed, and work on the thoughts that keep you awake ("if I don't sleep tonight, tomorrow is ruined"). But sleep restriction is the part that moves the numbers fastest.

Why does spending less time in bed help you sleep?

Three things happen at once.

1. Sleep pressure builds. The longer you are awake, the stronger the drive to sleep becomes. Going to bed later, with no naps, means you arrive in bed with more of it. You fall asleep faster and wake less.
2. Broken sleep joins up. Instead of 5.5 hours of sleep scattered across 8.5 hours, you get close to 5.5 hours in one piece. Fewer wakes in the night means fewer chances to start worrying.
3. Bed stops meaning "awake". After months of lying awake, the bed itself can become a cue for alertness. Many people with insomnia notice they doze on the sofa, then feel wide awake the moment they lie down in bed. When most of your time in bed is spent asleep, that link weakens.

A fixed wake-up time adds a fourth effect: your body clock gets one strong, daily anchor, so sleepiness starts to arrive at the same time each evening.

Does sleep restriction work? What the research says

The evidence is solid for the short term and growing for the long term.

• A 2021 meta-analysis in Sleep Medicine Reviews (Maurer and colleagues) pooled 8 randomized trials of sleep restriction against minimal or no treatment. It found large improvements in insomnia severity, sleep efficiency, time taken to fall asleep and time awake during the night. Total sleep time at the end of treatment did not change. The authors noted that most trials had a high risk of bias and few measured results past 3 months.
• The HABIT trial, published in The Lancet in 2023 (Kyle and colleagues), tested sleep restriction delivered by practice nurses in UK primary care in 642 adults with insomnia, in four short sessions. Insomnia severity improved more than with sleep advice alone at 3, 6 and 12 months.
• Guidelines. The American College of Physicians (2016) recommends CBT-I as the first treatment for chronic insomnia in adults, before sleeping pills. The American Academy of Sleep Medicine's 2021 guideline strongly recommends multi-part CBT-I and makes a conditional recommendation for sleep restriction on its own. The same guideline advises against "sleep hygiene" tips as a treatment by themselves.

In plain words: total sleep doesn't jump at first. What changes is that sleep gets deeper and less broken, and nights feel less like a battle. Total sleep tends to recover later, as the window is widened.

Who should not do sleep restriction alone

Shortening sleep on purpose is safe for most adults with ordinary insomnia, but not for everyone. Talk to a doctor first, and do it only with a clinician's guidance, if any of these apply:

• Bipolar disorder: sleep loss can trigger mania.
• Epilepsy or any seizure disorder: sleep loss can lower the seizure threshold.
• Loud snoring, gasping or pauses in breathing at night, or diagnosed sleep apnea that isn't treated. Daytime sleepiness from apnea plus restriction can be dangerous.
• You drive for a living, operate machinery or do safety-critical work.
• Pregnancy, a high risk of falls, or a medical condition your doctor says needs full sleep.
• You work night or rotating shifts (the method assumes one fixed wake time).
• You take medicines that make you drowsy, including sleeping pills: don't stop or change them without your prescriber.

And for everyone: if you feel sleepy, don't drive. The first two weeks are when sleepiness is highest. A 2014 study in Sleep (Kyle and colleagues) measured it: in the early weeks, people's measured sleep time went down and their daytime sleepiness and slower reaction times went up, before improving.

How to do sleep restriction: the step-by-step protocol

This is the common form used in CBT-I manuals. Clinicians vary the numbers a little; where they differ, we say so.

Step 1: Keep a sleep diary for 1 to 2 weeks

Each morning, write down four times: when you got into bed, roughly how long it took to fall asleep, how long you were awake in the night in total, and when you got out of bed. Estimates are fine. Don't watch the clock at night to get them; a rough guess in the morning is what the method uses.

From each night, work out two numbers:

• Time in bed: from getting in to getting out.
• Time asleep: time in bed, minus the time to fall asleep, minus the time awake in the night.

Then take the average of each across the week.

Step 2: Work out your sleep efficiency

Sleep efficiency is the share of time in bed you spend asleep: time asleep ÷ time in bed × 100. Good sleepers usually sit around 85% or higher. Someone asleep 5 hours 30 minutes out of 8 hours 30 minutes in bed has an efficiency of 65%.

Step 3: Set your first sleep window

Your first window equals your average time asleep. Some clinicians add 30 minutes to make the first week easier. Either way, there is a floor: most protocols never go below 5 hours, and many use 5.5 or 6 hours for older adults. If your average is under the floor, use the floor.

Step 4: Pick one wake-up time and keep it 7 days a week

Choose the time you need to be up on your earliest day, and keep it on weekends too. Set an alarm and get out of bed when it goes off, however the night went. This is the anchor of the whole method.

Step 5: Count back to your bedtime

Your earliest bedtime is the wake-up time minus the window. With a 6:30 a.m. wake-up and a 5-hour 45-minute window, the earliest you go to bed is 12:45 a.m. "Earliest" is the key word: if you aren't sleepy at 12:45, stay up until you are.

Step 6: Follow the rules inside the window

• Go to bed only when you are sleepy (eyes heavy, head nodding), not just tired.
• If you are awake for what feels like about 20 minutes, get up, go to another dim room and do something quiet until you feel sleepy again. Then go back to bed. Don't clock-watch to time it. Our guide on what to do when you can't fall asleep has a list of things to do out of bed.
• No naps, and no dozing on the sofa in the evening. They spend the sleep pressure you are building.
• Use the bed for sleep and sex only: no scrolling, TV or work in bed.
• Plan the hours before bedtime. Going to bed at 12:45 a.m. means a long evening; fill it with something you like, in good light, off the bed.

Step 7: Adjust the window once a week

Keep the diary going. At the end of each week, work out the week's average sleep efficiency and change the window by the rule in the table below. Change bedtime, never the wake-up time.

Sleep restriction rules: when to add or cut time

These are the cut-offs in a widely used CBT-I session guide (Perlis and colleagues). Others use 85% as the point to add time, and 80% as the point to cut, especially for adults over 60. Some clinicians also add time if you are sleeping well but still feel very sleepy in the day. Keep going until adding time no longer raises your total sleep, or until you feel rested in the day: that window is the one that fits you.

A worked example, week by week

Here is how the numbers move for someone whose diary showed 8 hours 30 minutes in bed (10:00 p.m. to 6:30 a.m.) and 5 hours 45 minutes asleep, an efficiency of 68%. It is an example of how the rules work, not a forecast for you: real weeks are bumpier.

Notice two things. In week 1, time asleep fell, which matches what studies see: the early weeks cost some sleep. And the gains are slow: 15 minutes a week. After six weeks this person sleeps about the same total as before, but in one piece, falling asleep quickly, with the 3 a.m. hour of lying awake gone. Widening continues from there.

What to expect in the first weeks

• Nights 1 to 3: staying up until your bedtime is the hardest part. Many people feel very sleepy in the evening and wonder why that never happened before.
• Week 1 to 2: more daytime sleepiness, sometimes irritability and headaches. You usually fall asleep faster and wake less. Keep the wake-up time even after a bad night: that's what makes the next night better.
• Weeks 3 to 4: efficiency climbs above 85%, the window starts to widen and daytime sleepiness eases.
• Weeks 5 to 8: the window keeps growing until it stops adding sleep. Most CBT-I programs run around 6 to 8 weeks.

When sleep restriction isn't working: fixes for common problems

Sleep compression: the gentler version

If a sudden cut sounds like too much, or you are older or need to be alert at work, there is a slower way. In sleep compression, you shrink the window gradually, often by 15 to 30 minutes a week, from your current time in bed down toward your time asleep, instead of all at once. It takes longer but causes less daytime sleepiness. Many clinicians use it for older adults.

With the example above: instead of jumping from 8 hours 30 minutes to 5 hours 45 minutes, you'd go 8 h 30, then 8 h, 7 h 30, 7 h, 6 h 30, stopping once efficiency reaches about 85%.

Sleep restriction is not sleep deprivation

The names sound alike, and people mix them up. Sleep deprivation means keeping someone awake. Sleep restriction keeps the sleep you already get and removes only the time you spend awake in bed. It does cause some short-term sleepiness, which is why the floor exists and why it's done for weeks, not months, before the window grows.

Doing it at home or with help

Many people do sleep restriction on their own from a book or a structured program. It is hard mainly because of the late evenings and the temptation to give up after a bad night, so a plan that tells you your exact bedtime each week, and keeps you on it, helps. Options:

• A CBT-I clinician (often a psychologist trained in behavioral sleep medicine). Ask your doctor for a referral, especially if any item in the "who should not do it alone" list applies.
• Digital CBT-I programs, some prescribed and some sold direct.
• A self-guided habit plan, like ours. To see where you'd start, our 2-minute sleep quiz works out your minutes awake in bed and a first sleep window from your own answers.

See a doctor if you have had trouble sleeping most nights for 3 months or more and it affects your days, if you snore loudly or stop breathing in your sleep, if your legs feel restless at night, if low mood or worry is part of it, or if you are thinking about stopping or starting any sleep medicine.

Frequently asked questions

How long does sleep restriction take to work?

Many people fall asleep faster within the first week or two, but feel more tired during those weeks. The window usually starts widening in weeks 2 to 4. Programs run about 6 to 8 weeks.

What is the minimum time in bed in sleep restriction?

Most protocols use a floor of 5 hours. Some use 5.5 or 6 hours, especially for older adults. If your average sleep is under the floor, start at the floor.

Can I nap during sleep restriction?

The standard method says no naps, because naps use up the sleep pressure you're building for the night. If you must nap for safety (before a drive, for example), keep it short and early in the afternoon, and tell your clinician.

Should I go to bed later or get up earlier?

Go to bed later. The wake-up time stays fixed. It's easier to stay up a little later than to get up earlier, and the fixed morning time is what sets your body clock.

Is sleep restriction safe for older adults?

It's used in older adults, often in the gentler compression form, with a higher floor and a lower efficiency cut-off. Falls risk and medicines matter more with age, so check with a doctor first.

Does sleep restriction help with waking at 3 a.m.?

Yes, that is one of the main things it targets: time awake in the middle of the night fell clearly in the trials. Our guide on waking up at 3 a.m. every night covers what to do in the moment.

Start tonight

Start the diary tomorrow morning and keep your wake-up time this weekend: that's the first step, and it costs nothing. To get your own starting window without the arithmetic, take the free sleep quiz. More on falling asleep, night waking and fixing your schedule: all our sleep guides.

Sources: Spielman AJ, Saskin P, Thorpy MJ, Treatment of chronic insomnia by restriction of time in bed, Sleep, 1987. Maurer LF et al., The clinical effects of sleep restriction therapy for insomnia: a meta-analysis of randomised controlled trials, Sleep Medicine Reviews, 2021. Kyle SD et al., Sleep restriction therapy for insomnia in primary care (HABIT), The Lancet, 2023. Kyle SD et al., Sleep restriction therapy for insomnia is associated with reduced objective total sleep time, increased daytime somnolence, and objectively impaired vigilance, Sleep, 2014. Qaseem A et al., Management of chronic insomnia disorder in adults, American College of Physicians guideline, Annals of Internal Medicine, 2016. Edinger JD et al., Behavioral and psychological treatments for chronic insomnia disorder in adults, American Academy of Sleep Medicine guideline, Journal of Clinical Sleep Medicine, 2021. Perlis ML et al., Cognitive Behavioral Treatment of Insomnia: A Session-by-Session Guide, Springer, 2005. Bootzin RR, Stimulus control treatment for insomnia, 1972. This guide is education, not medical advice.