How Long Before Bed Should You Stop Drinking Coffee? (2023 Meta-Analysis)
Share
About 8.8 hours for a standard coffee, and 13.2 hours for a pre-workout scoop. Those are the cutoffs a 2023 systematic review of 24 studies landed on if you want to avoid losing total sleep time. Six hours, the number most people have in their head, isn't enough: a randomized trial found that a 400 mg dose taken 6 hours before bed still cut measured sleep by over an hour, and the people losing that hour couldn't feel it happening.
Reviewed and published by Danny Lehr, CrossFit Level 3 Trainer and USA Weightlifting coach. October 2026.
Key findings at a glance
- To protect total sleep time, the 2023 meta-analysis recommends coffee (107 mg per 250 mL) at least 8.8 hours before bed, and a standard pre-workout serve (217.5 mg) at least 13.2 hours before bed.
- Pooled across 24 studies, caffeine cut total sleep time by 45 minutes, dropped sleep efficiency by 7%, added 9 minutes to falling asleep and 12 minutes of waking in the night.
- Deep sleep fell by 11.4 minutes and light sleep rose by 6.1 minutes.
- In the 2013 randomized trial, 400 mg at bedtime, 3 hours before bed, or 6 hours before bed each significantly reduced total sleep time versus placebo (p < 0.05 for all three), by 1.1 to 1.2 hours.
- Those three timings were not significantly different from each other. Six hours out was not measurably gentler than a dose at bedtime.
- At 6 hours out, the participants' own sleep diaries missed it. Their reported 41 minute loss was not statistically significant (p = 0.08).
How many hours before bed should you stop drinking caffeine?
Roughly 9 hours for a cup of coffee, and over 13 hours for a pre-workout scoop. The 2023 review pooled 24 studies specifically to answer this, and it reports the cutoff as dose-dependent: the bigger the hit, the further back it has to go. For a 10 p.m. bedtime that's a last coffee around 1 p.m., and a pre-workout dose before about 9 a.m.
"To avoid reductions in total sleep time, coffee (107 mg per 250 mL) should be consumed at least 8.8 h prior to bedtime and a standard serve of pre-workout supplement (217.5 mg) should be consumed at least 13.2 h prior to bedtime."
Isn't six hours the usual advice?
It is, and the trial that's usually cited for it doesn't actually support stopping there. The 2013 study tested 400 mg at 0, 3 and 6 hours before bed. All three significantly cut sleep, and crucially the three timings were statistically indistinguishable from each other. Six hours out behaved about like swallowing it as you got into bed. The authors framed 6 hours as a minimum, not a target, and they never tested anything further out.
"Even at 6 hours, caffeine reduced sleep by more than 1 hour."
Why can't you feel caffeine disrupting your sleep?
Because at 6 hours out, only the monitor caught it. The participants' sleep diaries showed a 41 minute loss and a more than doubled time to fall asleep, and neither reached significance (p = 0.08 and p = 0.06). The researchers flag that gap as one of the important results of the study, and as a likely reason people ignore caffeine timing advice. You are not a reliable witness to your own sleep.
What are the numbers?
| Measure | Finding | Statistically significant? |
|---|---|---|
| Cutoff for coffee, 107 mg (2023 review) | 8.8 h before bed | Modelled recommendation, not a tested timepoint |
| Cutoff for pre-workout, 217.5 mg (2023 review) | 13.2 h before bed | Modelled recommendation, not a tested timepoint |
| Total sleep time, pooled (2023 review) | Down 45 min | Pooled estimate; abstract reports no per-outcome test |
| Sleep efficiency, pooled (2023 review) | Down 7% | Pooled estimate; abstract reports no per-outcome test |
| Deep sleep, pooled (2023 review) | Down 11.4 min | Pooled estimate; abstract reports no per-outcome test |
| Total sleep time, caffeine 6 h before bed vs placebo (2013 trial) | Down 1.1 to 1.2 h | Yes, p < 0.05 |
| Wake time during sleep, 6 h before bed vs placebo (2013) | Up 8 min | Yes |
| Time to fall asleep, 6 h before bed vs placebo (2013) | Up 24.1 min | No |
| REM sleep (2013) | No change | No |
| Self-reported sleep loss at 6 h out (2013) | Down 41 min | No, p = 0.08 |
| 6 h vs 3 h vs bedtime (2013) | No difference detected | No |
The 1.1 to 1.2 hour figure is the range the 2013 paper reports across all three timings. It does not publish a separate number for the 6 hour condition, though the discussion states the 6 hour reduction exceeded 1 hour.
Is 400 mg of caffeine the same as a cup of coffee?
Not close. In the 2013 trial, 400 mg arrived as a pill, all at once, and the participants averaged 115 mg of caffeine a day at baseline. A standard 8 oz drip coffee runs roughly 95 to 165 mg. So the dose under test looks far more like a scoop of pre-workout or a large energy drink than the mug on your desk. That's also why the two cutoffs differ so much: dose drives the clock.
What should you do if you train at night?
Front-load the caffeine, hard. A 5 p.m. pre-workout scoop before a 10 p.m. bedtime is more than eight hours short of what the 2023 review recommends for that dose. If you train in the evening, either take the caffeine in the morning and let the session run on its own, or accept that you're trading sleep for it and decide that on purpose. Nobody here is telling you to drink less: morning and early afternoon coffee sit comfortably inside every cutoff above. If you're going to put the caffeine early, the cup should be worth drinking: PR Blend is roasted to order and carries about 50% more caffeine than a standard cup, which is what you want when it's doing work before a session instead of propping up a night. The rest of the coffee lineup covers the cup you drink for the taste, and the Coffee of the Month annual plan keeps fresh bags arriving so the morning cup is never the stale one.
What this does not prove
- The 8.8 and 13.2 hour figures are recommended cutoffs derived from the reviewers' modelling, not timepoints anyone randomly assigned and tested. Treat them as the best current guidance, not a measured result.
- The 2023 abstract reports pooled effect sizes without per-outcome significance tests or confidence intervals, so those rows above are estimates rather than confirmed effects.
- The 2013 trial is 12 people. The authors name the small sample as a limitation and say it likely cost them the power to detect caffeine effects by self-report.
- No coffee was used in the 2013 trial. The dose was a 400 mg pill, so it says nothing directly about a specific brew.
- Its subjects were light-to-moderate caffeine users (115 mg a day) and young to middle-aged. Generalizing to heavy users, caffeine-naive people, or older adults needs more work.
- Sleep there was measured with a validated consumer EEG headband, not full polysomnography. The authors call the sleep-stage results preliminary and note the device can undercount wake epochs.
- That trial was funded by an investigator-initiated grant from Zeo Inc., the company that made the sleep monitor used.
- Neither paper measured training performance, strength, or body composition. These are sleep studies.
FAQ
How long before bed should I stop drinking coffee?
About 8.8 hours for a standard 107 mg coffee, according to a 2023 meta-analysis of 24 studies. For a 10 p.m. bedtime that's a last cup around 1 p.m.
How long before bed should I stop taking pre-workout?
About 13.2 hours for a standard 217.5 mg serve. The dose is roughly double a coffee, so the clearance window is substantially longer. For a 10 p.m. bedtime that's before about 9 a.m.
Is six hours before bed enough?
Probably not. In a randomized trial, 400 mg taken 6 hours before bed still significantly cut total sleep time, and was statistically indistinguishable from taking the same dose at bedtime.
Why can't I feel caffeine affecting my sleep?
Self-report misses it. At 6 hours out the monitor detected the lost sleep but the participants' own diaries did not reach significance. The authors treat that blind spot as a key result.
Should I quit coffee to sleep better?
Neither study gives a reason to. Neither tested removing coffee, and neither tested morning coffee. What they tested was timing and dose, which are the parts you can actually move.
Sources
- Gardiner C, Weakley J, Burke LM, Roach GD, Sargent C, Maniar N, Townshend A, Halson SL. The effect of caffeine on subsequent sleep: A systematic review and meta-analysis. Sleep Medicine Reviews. 2023;69:101764. PubMed abstract
- Drake C, Roehrs T, Shambroom J, Roth T. Caffeine effects on sleep taken 0, 3, or 6 hours before going to bed. Journal of Clinical Sleep Medicine. 2013;9(11):1195-1200. PubMed abstract
- McCusker RR, Goldberger BA, Cone EJ. Caffeine content of specialty coffees. Journal of Analytical Toxicology. 2003;27(7):520-522. PubMed abstract
This is population level research, not medical advice.