Caffeine Half-Life: Why Your 3 pm Coffee Is Still Working

A cup of coffee doesn't finish when you finish drinking it. Caffeine's average half-life in healthy adults is roughly five hours — the US FDA puts it at 4 to 6. A 200 mg coffee at 3 pm leaves about 100 mg circulating at 8 pm and close to 50 mg at midnight. That is not a trace amount; it is half a strong espresso acting on your central nervous system while you are trying to sleep.

The five-hour rule, and what it actually means

Half-life is the time your body needs to cut a substance's blood concentration in half. Caffeine is cleared mainly by the liver enzyme CYP1A2. Five hours is the midpoint, not a countdown to zero. After two half-lives (about 10 hours), a quarter of the dose remains. After three, an eighth. To clear roughly 97% of a dose you need five half-lives — about 25 hours.

Run the clock on a 3 pm pour. A 12 oz drip coffee carries around 90–140 mg depending on roast and brewing; a large cold brew can exceed 200 mg. At midnight, nine hours later, you are one and a half half-lives in with more than a quarter of that dose still in your bloodstream. Energy drinks and pre-workout powders push 150–300 mg per serving, so a single can at 3 pm behaves like two coffees.

The range matters as much as the average. Individual half-lives run from roughly 2.5 to 10 hours, according to pharmacokinetic data summarized by the NIH and FDA. Hormonal contraceptives can double clearance time, pregnancy extends it further, and smoking shortens it. Two people ordering the same flat white are not in the same chemical position by evening.

What caffeine does while you sleep

Caffeine doesn't create energy. It blocks adenosine, the molecule that accumulates through the day and signals your brain that sleep pressure is rising. Blocking the receptor masks the signal; it doesn't drain the tank. When caffeine fades, that built-up adenosine can land at once — the afternoon crash many drinkers blame on the coffee itself.

Inside the night, the effect is measurable. In a randomized trial by Drake and colleagues published in the Journal of Clinical Sleep Medicine, 400 mg taken zero, three or six hours before bed reduced objectively measured total sleep time by more than an hour — even in the 6 pm group, who had half a day of clearance. Sleep onset slowed, wake time grew, and deep sleep was cut. The participants were regular caffeine users; habit did not protect them.

Pooling the trial literature, a 2023 meta-analysis in Sleep Medicine Reviews found caffeine reduced total sleep time by roughly 45 minutes and sleep efficiency by about 7%, even when people reported falling asleep normally. Objective disruption tends to arrive before subjective insomnia does.

Why it matters

Translate that into a week. If afternoon caffeine shallows your sleep by 30–60 minutes a night, you start the next day under-slept, which raises perceived fatigue and nudges you toward another afternoon cup. The half-life math means that cup repeats the cycle. Caffeine is not the cause of everyone's insomnia, but for a large share of poor sleepers it is a mechanical, predictable contributor — and one of the few you can fix without medication.

Deep sleep takes the hit first. Slow-wave sleep is when the body consolidates physical recovery and, in research models, clears metabolic waste from the brain. Trading it for a 3 pm coffee is a bad exchange rate: you pay with your most restorative sleep for alertness you may not need after 8 pm.

Opinion: for beginners in sleep hygiene, the caffeine cutoff is the single highest-leverage change. It costs nothing, works within days rather than months, and unlike most gadgets, it targets a documented mechanism.

The honest caveat

Here is the legitimate objection: averages hide variation. Fast CYP1A2 metabolizers can clear a dose in two to three hours and may sleep fine after an evening espresso. Plenty of people report decades of 9 pm coffee with no complaints, and they argue population math shouldn't dictate personal habit.

That objection has real force — with two limits. First, subjective tolerance routinely outpaces objective sleep architecture: trials consistently detect reduced deep sleep and efficiency in regular users who do not feel impaired. Second, the variation cuts both ways; if you are a slow metabolizer, the standard advice is already too generous for you. The practical answer is not to debate averages but to test yourself. Keep your usual pattern for a week, then move all caffeine before 2 pm for two weeks and compare how fast you fall asleep and how you feel before that first coffee. Call it what it is: an n-of-1 experiment, not a guarantee.

Practical cutoffs that respect the half-life

• Baseline rule: last dose 8–10 hours before bed — roughly 1–3 pm for an 11 pm bedtime. A 3 pm coffee starts to look expensive under that math.
• Count every source: coffee, tea, cola, energy drinks, pre-workout, dark chocolate. Decaf still carries 2–15 mg per cup.
• If pregnant, on hormonal contraceptives, or unusually sensitive, extend the cutoff or halve the dose instead of trusting the average.
• Taper rather than quit cold turkey. Dropping a heavy habit abruptly can trigger headaches and irritability within a day; NIH's NCCIH notes these withdrawal symptoms typically pass within about a week.
• Stay at or under 400 mg per day, the FDA's ceiling for healthy adults.

The bottom line

Five hours is a midpoint, not a finish line. A 3 pm coffee at midnight is not folklore — it is a quarter to half a dose still on board, measurably shaving sleep you need. Move your cutoff earlier for two weeks and let your own sleep break the tie.