by Strong Mornings See our plans

Does CBT really work for insomnia? What the research shows

Does CBT really work for insomnia? Yes, on average, and it is one of the few sleep questions with a firm answer. In a 2015 meta-analysis of trials (Trauer and colleagues, Annals of Internal Medicine), CBT-I shortened the time to fall asleep by about 19 minutes, cut time awake in the night by about 26 minutes and raised sleep efficiency by about 10 percentage points. It added very little total sleep time, about 8 minutes. The American College of Physicians (2016) and the American Academy of Sleep Medicine (2021) both recommend it as the first treatment for chronic insomnia.

This guide is not medical advice. Below: what "work" means in the trials, a table of the numbers, who it works less well for, why it fails for some people, how fast to expect changes, how it compares with pills, and how to give it a fair try. To see where a sleep window would start for you, the free 2-minute sleep quiz uses your own numbers.

Does CBT really work for insomnia? What the studies found

MeasureAverage change with CBT-IWhat it means for you
Time to fall asleepAbout 19 minutes shorterLess lying awake at lights out
Time awake after falling asleepAbout 26 minutes shorterShorter or fewer wake-ups
Sleep efficiencyAbout 10 points higherMore of your time in bed is spent asleep
Total sleep timeAbout 8 minutes longerSleep gets better, not much longer

These are averages across trials, so some people did better and some did worse. The point is that CBT-I changes how sleep feels and how long you lie awake. It does not turn a 6-hour sleeper into an 8-hour one. A small figure for total sleep time is why some people say it "does not work": they were hoping for hours and got a better night of the same length. Our CBT for insomnia techniques guide explains which step makes which change.

Do the guidelines back it?

Yes. The American College of Physicians recommended CBT-I as the initial treatment for chronic insomnia in adults in 2016, with medicine considered only if it does not work or is not an option. The American Academy of Sleep Medicine published a clinical practice guideline on behavioral and psychological treatments in 2021 that recommends multicomponent CBT-I. Compare it with medicine in CBT-I vs medication.

Does online or app-based CBT work too?

It has been tested. A 2016 meta-analysis of internet-delivered CBT-I (Zachariae and colleagues) found improvements in insomnia severity and sleep measures against control groups. A trial of Sleepio in 3,755 university students found benefits for insomnia and mental health. Digital programs may do a little less than a clinician in some trials, and they suit people with straightforward insomnia best. Options are in CBT-I courses online and CBT for insomnia apps.

Why it seems not to work for some people

  1. They stopped in week 2. A short sleep window makes you more tired before it makes you sleep better. This is expected and is the point where most people quit.
  2. They did the habits and skipped the window. Sleep hygiene alone does not do what the full method does. See CBT-I vs sleep hygiene.
  3. The wake time kept moving. A fixed wake time seven days a week is the anchor. Sleeping in on weekends undoes much of it.
  4. The window was too long. Set from the wish to sleep eight hours rather than the diary. See the sleep diary template for the sums.
  5. Another problem was in the way. Untreated sleep apnea, restless legs, a medicine, alcohol, pain or depression can all keep sleep broken whatever the method.
  6. They stayed in bed when awake. Without getting up, the bed keeps its link to wakefulness. See stimulus control.

How fast does it work?

Programs run roughly 4 to 9 weeks. Many people notice the diary improving in the second to fourth week, and the full effect builds over the course. The first week or two can feel worse, with more daytime tiredness while the window is tight. Judge it at week 6 to 8 by your diary: time to fall asleep, time awake at night and sleep efficiency, not by how you feel on one morning.

What counts as "working"? Measure it yourself

You do not need a trial to know if CBT-I is working for you; your diary tells you. Compare the average of your first two diary weeks with the average of weeks 6 to 8 on three numbers: minutes to fall asleep, minutes awake at night, and sleep efficiency (time asleep divided by time in bed). A fall of 15 to 20 minutes in either of the first two, or a rise of about 10 points in efficiency, matches what trials found on average. The sleep diary template shows how to do the sums.

Diary measureWeeks 1–2 (example)Weeks 6–8 (example)Reading
Minutes to fall asleep5025Down 25: better than the trial average
Minutes awake at night6035Down 25: in line with the average
Sleep efficiency65%82%Up 17 points: a good response
Total sleep time5 h 10 min5 h 40 minUp 30 minutes: more than the average gain

These example numbers are made up to show the sums, not results from any person or trial.

If it did not work for you, what next?

First check that you actually ran the whole method for six or more weeks: diary, fixed wake time, a window set from the diary, getting out of bed when awake. If you did, a clinician is the next step, because something may be in the way. A sleep doctor can test for apnea or leg movements, a CBT-I specialist can adjust the plan (see insomnia therapist near me), and a doctor can review medicines, alcohol and mood. A second round of CBT-I with a clinician often works where the first, self-guided one stalled, but we do not have a trial figure to quote for that.

Who it may work less well for

CBT-I vs sleeping pills: which has better evidence?

Pills usually work fast and can help in a crisis, but the effect tends to fade once you stop, and the guidelines suggest weighing side effects. CBT-I takes weeks and its gains tend to last after the program ends in many trials. Many people use both for a while, with a doctor, and then taper. Do not stop a prescription on your own. For the comparison, see CBT-I vs medication.

How to give it a fair try

  1. Keep a sleep diary for two weeks before you change anything.
  2. Set one fixed wake time you can keep every day.
  3. Use a sleep window from your diary and hold it for at least four weeks.
  4. Get out of bed after about 20 minutes awake and return only when sleepy.
  5. Write worries down earlier in the evening, not in bed.
  6. Review the diary weekly and adjust the window by the rules, not by mood.
  7. If nothing has moved at week 6, see a doctor or a CBT-I therapist.

Frequently asked questions

Does CBT really work for insomnia long term?

Many trials found the gains held after treatment ended, which is why guidelines prefer it to a pill as first choice. Long-term results vary by person, and a relapse plan helps.

How many people does CBT-I help?

We have not quoted a single percentage because figures differ by trial and by how "response" is defined. Meta-analyses show average gains in sleep onset, time awake at night and efficiency.

Is CBT-I better than sleep hygiene?

Sleep hygiene is one part of the picture. The full method adds a sleep window and stimulus control, which carry most of the effect.

Does CBT-I work for insomnia with anxiety?

It can, and it can lower anxiety about sleep. If anxiety is severe, a clinician can treat both. See sleep anxiety.

Can I try CBT-I on my own?

Yes, with a workbook, an app or a course, unless one of the cautions above applies. See our CBT for insomnia workbook guide.

Start with the free part

The first step is the same whichever way you go: a diary and a fixed wake time. The free sleep quiz gives you a starting window in two minutes, and all sleep guides lists the rest.

Sources: Trauer JM et al., Cognitive behavioral therapy for chronic insomnia: a systematic review and meta-analysis, Annals of Internal Medicine, 2015 (figures are pooled averages, rounded). Qaseem A et al., ACP clinical practice guideline on chronic insomnia, Annals of Internal Medicine, 2016. Edinger JD et al., AASM clinical practice guideline on behavioral and psychological treatments for chronic insomnia, Journal of Clinical Sleep Medicine, 2021. Zachariae R et al., internet-delivered CBT for insomnia meta-analysis, Sleep Medicine Reviews, 2016. Espie CA et al., Sleepio in university students, The Lancet Psychiatry, 2017. Read each paper before quoting its numbers.