Caffeine and sleep architecture

This page is a reference summary of published evidence and regulatory documents. It is not a substitute for individual medical advice, and it is not a recommendation to take or avoid any medicine.

Medically reviewed by Dr. Sofia Lindqvist, · Last reviewed 11 August 2026

Caffeine measurably alters sleep architecture even in people who do not notice any effect on how quickly or well they fall asleep. Polysomnography studies show reduced slow-wave sleep, increased sleep latency and wake after sleep onset (WASO), and reduced total sleep time following afternoon or evening caffeine intake.

Changes to sleep architecture

Objective sleep recordings after caffeine intake show reduced slow-wave sleep duration and reduced slow-wave activity (SWA), the EEG marker most closely tied to sleep depth and homeostatic recovery. Sleep latency (time to fall asleep) and WASO (time awake after initially falling asleep) both increase, and total sleep time is reduced, in some cases without the person reporting subjectively worse sleep.

Drake et al. 2013

A controlled study by Drake and colleagues gave participants 400mg of caffeine at 0, 3, or 6 hours before bedtime.

Timing before bed	Effect on total sleep time vs placebo
0 hours	Significant reduction
3 hours	Significant reduction
6 hours	Reduction of approximately 1 hour

Caffeine taken 6 hours before bedtime still reduced total sleep time by about an hour, despite being well outside the period most people associate with a bedtime stimulant.

Burke et al. 2015: circadian phase delay

Burke and colleagues found that caffeine taken in the evening, roughly 3 hours before habitual bedtime, produced a measurable delay in the circadian phase of melatonin onset, an effect in the same direction as, though smaller than, exposure to evening light. This indicates caffeine affects the timing of the circadian clock, not only sleep depth and continuity.

Half-life and timing

Caffeine’s average half-life is approximately 5 hours, so a substantial fraction of an afternoon or evening dose is still present in the body at typical bedtimes. This pharmacokinetic fact is consistent with sleep-architecture effects being measurable many hours after intake, as observed in the Drake 2013 data.

Frequently asked questions

Does caffeine affect sleep even if I feel fine falling asleep?

Yes. Studies using polysomnography show reduced slow-wave sleep, increased wake after sleep onset, and reduced total sleep time even when subjects do not report noticing an effect.

How much sleep did caffeine cost in the Drake 2013 study?

400mg of caffeine taken 6 hours before bedtime reduced total sleep time by about 1 hour compared with placebo.

Can caffeine shift the body clock?

Yes. Burke et al. (2015) found that evening caffeine intake produced a measurable phase delay of the circadian clock, comparable in direction to the effect of dim light exposure.

What does caffeine do to slow-wave sleep?

It reduces slow-wave sleep duration and slow-wave activity (SWA), a marker of sleep depth and homeostatic sleep pressure recovery.

Why does caffeine's half-life matter for sleep?

With an average half-life of about 5 hours, a substantial fraction of an afternoon or evening dose remains in the body at bedtime, which is consistent with the sleep-architecture effects measured hours after intake.