A Bedtime Routine for Adults That Works, Step by Step

A woman in a navy dressing gown sits on the edge of a made bed in daylight, one hand resting on a paperback beside a yellow hot-water bottle.
In this article
  1. Does a bedtime routine help adults sleep, and when should it start?
  2. Step 1: when should the last coffee be?
  3. Step 2: is it bad to exercise in the evening?
  4. Step 3: does a warm bath or shower help you sleep?
  5. Step 4: dim the room, not just the phone
  6. Step 5: what should the hot drink be?
  7. Step 6: write tomorrow down and set a lights-out time
  8. Step 7: the same wake time, including weekends
  9. What is the 10-3-2-1 rule, and does it hold up?
  10. What if you are still awake, or the routine does not fix it?
  11. What we don't know
  12. Frequently asked questions

A bedtime routine for adults works when it protects the same wake time seven days a week, lowers light and stimulation for the last 90 minutes, and keeps bed for sleep rather than for lying awake. In 60,977 UK Biobank adults wearing wrist monitors, regular sleepers had a 20% to 48% lower risk of dying during follow-up than the least regular fifth, and regularity predicted mortality more strongly than hours slept.1 Each step carries its trial; the timed routine is practical guidance. The wider picture is in the Wellgard sleep guide.

Sources checked on 30 October 2026.

Does a bedtime routine help adults sleep, and when should it start?

It helps, within limits. Across 15 studies, sleep hygiene education alone improved diary-recorded sleep efficiency by about 5% and the Pittsburgh Sleep Quality Index by two points, only on self-report; cognitive behavioural therapy for insomnia (CBT-I) added about 8% more.2 The NHS says: "Insomnia usually gets better by changing your sleeping habits."3

The NHS says to "relax at least 1 hour before bed – for example, take a bath or read a book".3 A 2022 expert consensus on indoor lighting recommends, in its full text, at most 10 lux of melanopic equivalent daylight illuminance from three hours before bed: lamps, not ceiling lights.4 Ninety minutes sits between the two. As practical guidance, for an 11pm bedtime and a 7am alarm:

  1. By 2pm: last caffeine (step 1).
  2. By 9.30pm: work closed, exercise finished, lamps on and ceiling lights off (steps 2 and 4).
  3. 9.30pm to 10pm: warm bath or shower (step 3).
  4. 10pm: hot drink, tomorrow's list, phone charging outside the bedroom (steps 5 and 6).
  5. 10pm to 11pm: a paper book in lamp light; bed when sleepy, not by the clock.
  6. 7am, every day: alarm, curtains open, up (step 7).

Step 1: when should the last coffee be?

In a 2023 meta-analysis of 24 studies, caffeine cut total sleep by 45 minutes and sleep efficiency by 7%, added 9 minutes to falling asleep and 12 minutes awake in the night, and took 11.4 minutes off deep sleep. To protect total sleep, a 107 mg coffee needed to be at least 8.8 hours before bed, a 217.5 mg pre-workout serving 13.2 hours.5 That is a modelled cut-off, and sensitivity varies widely. The NHS floor: "do not smoke or drink alcohol, tea or coffee at least 6 hours before going to bed".3 Cup sizes and half-life: the caffeine and sleep page.

Step 2: is it bad to exercise in the evening?

Not for most people. A 2019 meta-analysis of 23 studies in healthy adults found a single evening session did not harm sleep, adding 7.7 minutes before the first REM period and 1.3 percentage points of slow-wave sleep; only vigorous exercise ending an hour or less before bed might impair falling asleep, total sleep and efficiency.6 The NHS gives the cautious version: "do not exercise at least 4 hours before bed".3

Step 3: does a warm bath or shower help you sleep?

Yes, if it is timed. A 2019 review found 17 studies of warm bathing or showering before bed, 13 poolable. Water at 40 to 42.5 °C went with better self-rated sleep quality and higher sleep efficiency, and taken 1 to 2 hours before bed, for as little as 10 minutes, it significantly shortened the time to fall asleep, which the authors attribute to a faster fall in core temperature.7 The research base is thin; a bath at lights-out misses the window.

Step 4: dim the room, not just the phone

In 116 healthy volunteers aged 18 to 30, room light under 200 lux in the eight hours before bed, against dim light under 3 lux, delayed melatonin onset in 99% of individuals and shortened its duration by about 90 minutes.8 In a laboratory crossover study of 12 young adults, reading on a light-emitting e-reader before bed meant longer to fall asleep, less melatonin, a later body clock and lower next-morning alertness than a printed book.9 A phone's warm-screen night setting did not rescue it: in a randomised trial of 167 adults aged 18 to 24 using a phone for the hour before bed for seven nights, sleep did not differ significantly between the warm setting, the normal screen and no phone; only a post-hoc analysis of those already sleeping over 6.8 hours favoured no phone.10 The NHS: "do not watch television or use devices, like smartphones, right before going to bed, because the blue light makes you more awake".3 See tired but wired.

Step 5: what should the hot drink be?

A hot drink marks the wind-down, and the NHS Every Mind Matters page suggests one: "You could try a warm milky drink or herbal tea instead."11 Chamomile, lemon balm, tart cherry, warm milk and sleep powders each have their trials on the what to drink before bed page. If a supplement is part of your evening, magnesium's authorised wording is that it contributes to normal psychological function and to the reduction of tiredness and fatigue;16 what the trials found is on that page. One small mug, an hour before bed.

Step 6: write tomorrow down and set a lights-out time

The NHS suggests: "If you often worry in bed, try making a list of things you need to do the next day before you go to sleep."11 Across 15 randomised trials, CBT-I had moderate-to-large effects on worry (Hedges' g -0.41 to -0.71) but small, unreliable effects on rumination (g -0.13).12 For the racing-mind version, see how to switch off your mind at night.

Bedtime procrastination, going to bed later than intended when nothing stops you, was described in a 2014 survey of 177 adults and went with insufficient sleep beyond what demographics and self-regulation explained; the evidence is correlational, and the "revenge" prefix is social media's.13 The fix: a lights-out alarm, and the phone charging in another room.

Step 7: the same wake time, including weekends

Regularity, not duration, predicted mortality in the 60,977 UK Biobank adults.1 In 73,513 UK adults followed for eight years, catching up two or more hours at weekends was not associated with lower mortality (hazard ratio 1.17, 95% confidence interval 0.97 to 1.41, not significant) or cardiovascular disease (1.05, 0.94 to 1.18, not significant).14 The NHS: "wake up and get out of bed at the same time every day" and "do not sleep in after a bad night's sleep – stick to your regular sleeping hours instead".3 See the clocks going back and your sleep.

What is the 10-3-2-1 rule, and does it hold up?

No caffeine 10 hours before bed, no food or alcohol 3 hours before, no work 2 hours before, no screens 1 hour before. None of it comes from a trial (viral sleep rules and tricks), but the shape is right. The 10 is stricter than the NHS six and close to the modelled 8.8 hours.3,5 The 3 has no trial; the NHS says "do not eat a big meal late at night" and to avoid alcohol and nicotine "1 to 2 hours before bed"3,11 (why alcohol wakes you up at 3am). The 2 is unmeasured. The 1 matches the NHS advice to avoid phones, tablets and computers "for an hour before bed",11 though room lights matter too.8

What if you are still awake, or the routine does not fix it?

Do not lie there. The NHS says "only go to bed when you're sleepy", and Every Mind Matters adds that "if you're still awake after around 20 minutes, get up and sit somewhere comfortable" and "go back to bed when you feel sleepy".3,11 This is stimulus control. In a component network meta-analysis of 241 randomised trials with 31,452 adults, stimulus control was associated with remission (incremental odds ratio 1.43, 95% confidence interval 1.00 to 2.05) and with better sleep quality, efficiency and time to fall asleep; sleep hygiene education alone was not essential (1.01, 0.77 to 1.32), and relaxation procedures were potentially counterproductive (0.81, 0.64 to 1.02, not significant).15

If the routine has not helped, the NHS says to see a GP when "changing your sleeping habits has not helped your insomnia" or "you've had trouble sleeping for months"; a GP may offer cognitive behavioural therapy.3 More on when to see a GP about sleep.

What we don't know

No trial has tested a whole bedtime routine against none; each step has its own evidence. The sleep hygiene review found no study comparing it with sham or no treatment,2 the warm-bath findings rest on few studies with the best timing unclear,7 and the warm-screen phone trial found no difference across its full sample.10

Frequently asked questions

What should you never do right before bed?

The NHS insomnia page's don't list runs: alcohol, tea or coffee in the six hours before bed; a big meal late at night; exercise in the four hours before bed; television or devices right before bed; "do not nap during the day"; "do not drive when you feel sleepy"; and sleeping in after a bad night.3

Is it bad to eat before bed?

The NHS says "do not eat a big meal late at night", and Every Mind Matters that "eating large meals close to bedtime may make it harder to fall asleep".3,11 Neither sets a number of hours (foods that help you sleep).

What is revenge bedtime procrastination?

Going to bed later than intended when nothing stops you. In the 2014 survey of 177 adults that introduced the term, it went with reporting insufficient sleep, though the design cannot show cause.13 The lights-out alarm is aimed at it.

Sources

  1. Windred DP, et al. Sleep, 2024. Cohort, 60,977 adults; NIH grant. PubMed 37738616
  2. Chung KF, et al. Fam Pract, 2018. Meta-analysis, 15 studies; funding not stated. PubMed 29194467
  3. NHS. Insomnia. Reviewed 19 March 2024. nhs.uk/conditions/insomnia
  4. Brown TM, et al. PLoS Biol, 2022. Expert consensus; UK research council (BBSRC) workshop grant. PubMed 35298459
  5. Gardiner C, et al. Sleep Med Rev, 2023. Meta-analysis, 24 studies; funding not stated. PubMed 36870101
  6. Stutz J, et al. Sports Med, 2019. Meta-analysis, 23 studies; funding not stated. PubMed 30374942
  7. Haghayegh S, et al. Sleep Med Rev, 2019. Meta-analysis, 17 studies; funding not stated. PubMed 31102877
  8. Gooley JJ, et al. J Clin Endocrinol Metab, 2011. Laboratory analysis, 116 adults; NIH-funded. PubMed 21193540
  9. Chang AM, et al. Proc Natl Acad Sci USA, 2015. Crossover study, 12 adults; NIH-funded. PubMed 25535358
  10. Duraccio KM, et al. Sleep Health, 2021. Randomised trial, 167 adults; funding not stated. PubMed 33867308
  11. NHS Every Mind Matters. How to fall asleep faster and sleep better. Read 30 October 2026. nhs.uk/every-mind-matters
  12. Ballesio A, et al. Sleep Med Rev, 2021. Meta-analysis, 15 trials; funding not stated. PubMed 32992228
  13. Kroese FM, et al. Front Psychol, 2014. Survey, 177 adults; Dutch research council. PubMed 24994989
  14. Chaput JP, et al. Sleep, 2024. Cohort, 73,513 adults; funding not stated. PubMed 38895883
  15. Furukawa Y, et al. JAMA Psychiatry, 2024. Network meta-analysis, 241 trials, 31,452 adults; Japanese public grants. PubMed 38231522
  16. DHSC. Great Britain nutrition and health claims register. Updated 19 May 2026. gov.uk

Spotted something out of date? Email support@wellgard.co.uk with the page address and the sentence.