Coffee and Sleep: What Research Shows

Caffeine and sleep are on opposite sides of the same clock. Caffeine delays the biology that makes you tired; sleep does its work when that biology is allowed to run. The practical question isn't whether coffee affects sleep — it does — but how much, for how long, and at what times of day it starts to matter.

What caffeine actually does to sleep

Caffeine blocks adenosine, a neurotransmitter that builds up in your brain during waking hours and creates the pressure to sleep. Blocking adenosine is why coffee feels stimulating; it's also why coffee disrupts sleep, because the same pressure signal your body uses to initiate rest is muted.

The effect shows up in three measurable ways:

• Longer sleep onset. It takes longer to fall asleep after a caffeine dose.
• Reduced total sleep time. You wake up sooner or more often.
• Shifted sleep architecture. Less time in deep (slow-wave) sleep, more time in lighter stages. This is why you can technically "sleep eight hours" after a late coffee and still feel unrested.

A 2024 randomised crossover trial published in SLEEP directly tested this. Healthy adults were given either placebo, a typical caffeine dose, or a high dose in the morning, afternoon, or evening, and their subsequent sleep was measured. Every caffeine condition reduced sleep duration and quality compared to placebo, and the effect was worse the later in the day the caffeine was taken^[1].

The half-life question

Caffeine's half-life — the time for your body to clear half of a dose — sits at roughly four to six hours in healthy adults^[2]. This is the number behind the standard "no coffee after 2 PM" advice, and the arithmetic is useful:

Numbers assume a 5-hour half-life and an 11 PM bedtime; individual half-lives vary substantially.

Individual sensitivity is real and largely genetic

Some people drink an after-dinner espresso and sleep fine. Others notice a lunchtime cup at 10 PM. The difference is not willpower.

Two well-studied genetic variants explain a large fraction of the person-to-person spread:

• CYP1A2 — the liver enzyme that metabolises caffeine. Fast metabolisers clear caffeine roughly twice as quickly as slow metabolisers, so the same cup lingers half as long in their system^[2].
• ADORA2A — the adenosine A2A receptor gene. Variants in this gene predict how strongly caffeine will disrupt sleep. People with certain variants show measurably worse caffeine-induced sleep disruption at the same dose, independent of how fast they metabolise it^[2].

Practical implication: population averages predict individuals badly. If you sleep poorly and drink caffeine, cutting it back is worth an experiment even if "most people" tolerate afternoon coffee.

Practical rules that actually help

• Stop caffeine by early afternoon if sleep matters. The "8-hours-before-bed" rule is a safe default. If you go to bed at 11 PM, that means no caffeine after 3 PM.
• Count all caffeine, not just coffee. Tea, chocolate, cola, pre-workouts, some medications, and matcha lattes all count.
• Watch total daily dose, not just timing. A morning triple espresso can still be affecting sleep 12 hours later in slow metabolisers.
• Experiment with a week off. A one-week caffeine-free trial tells you more about your personal sleep sensitivity than any general guideline.
• Notice sleep quality, not just duration. You can "get eight hours" and still lose deep sleep to a late coffee.

What sleep-friendly coffee looks like

If you enjoy coffee but sleep matters:

• Front-load your intake. Two morning cups affect sleep less than one morning cup and one after-dinner cup.
• Switch to decaf after your cutoff. Decaf provides the ritual and flavor at 2–15 mg of caffeine per cup — usually not enough to meaningfully disturb sleep^[3].
• Consider a lower-caffeine drink for the afternoon. Green tea, black tea, or a half-caf blend all deliver noticeably less caffeine than a full cup of coffee.
• Skip the "one last coffee" habit. Late-evening caffeine has the strongest effect on sleep of any timing tested in the 2024 trial — including for people who don't feel it acutely.

The honest summary

Caffeine disrupts sleep more than most people notice. The effect is biggest at high doses and late timing, but even morning coffee has a measurable effect in controlled trials. Individual sensitivity varies widely and is largely genetic — the safest default is to stop by early afternoon and see whether your sleep feels different.