How Many Hours Before Bed Should You Stop Drinking Coffee?

The 8.8-hour rule you have seen quoted everywhere comes from a 2023 meta-analysis of 24 studies, which modeled the point at which a 107 mg cup of coffee stops measurably shortening sleep (see [2]). Pooled across those studies, caffeine cut total sleep time by 45 minutes and sleep efficiency by 7 percent. Real numbers, but averaged ones, and a crossover trial published in Sleep in 2025 shows why that matters. Twenty-three healthy men slept with in-home polysomnography after 100 mg or 400 mg of caffeine taken 12, 8, or 4 hours before bed. The 100 mg dose produced no significant effect on objective or subjective sleep at any of the three timings, including four hours out (see [1]).

The 400 mg arm is where it fell apart. Four hours before bed it cost about 51 minutes of total sleep, added 14 minutes to sleep onset, and dropped sleep efficiency by 9.5 percent. Eight hours out it still added 26 to 29 minutes of wake after sleep onset. Slow wave sleep, the part worth protecting, was down 21 minutes even when the dose landed a full 12 hours before bed (see [1]). So the honest answer is not a time, it is a dose crossed with a time. An afternoon coffee and a scoop of pre-workout are not the same experiment.

There is a second layer the sleep-stage summary misses. A 2025 Communications Biology study gave 40 adults aged 20 to 58 either 200 mg of caffeine or placebo and looked at the EEG itself. Caffeine pushed the sleeping brain toward higher complexity and closer to criticality, a more excitable and more awake-like regime, with the shift clearest in non-REM sleep (see [3]). During REM the effect was much larger in the 20 to 27 group than the 41 to 58 group, which the authors tie to age-related changes in adenosine receptors. You can sleep the same number of hours and not be doing the same thing with them.

Both trials are small, young, and in the 2025 case all male (see [1]), so any published cutoff is a starting hypothesis, not your number. The clean N=1 version: log the clock time and the milligrams for two weeks, alternate an early-cutoff block against your normal habit, and compare sleep efficiency and wake after sleep onset across the blocks instead of reacting to one bad night. If you track sleep in Wellness Project, the /sleep page lines those nights up against what you logged. For how the deficit accumulates when those nights stack, see our guide to sleep debt.