Caffeine’s half-life, drawn on your actual evening

The maths of a 15:00 coffee

A large coffee holds roughly 200 mg of caffeine. Caffeine's average half-life is about five hours: every five hours, your liver clears half of what remains. Drink it at 15:00 and the arithmetic is fixed. 100 mg at 20:00. 66 mg at 23:00. 50 mg at 01:00. Even at 07:00 the next morning, around 22 mg is still circulating — a quarter-espresso you never drank.

The curve below is not a metaphor. It is the same exponential decay your liver runs, drawn against a normal evening. Note what it never does: reach zero. Caffeine does not wear off; it dilutes. The felt effect fades much earlier than the chemical does, which is exactly why the tail is easy to ignore.

What 66 mg does at bedtime

Caffeine works by blocking adenosine, the molecule that builds through the day and produces sleep pressure. 66 mg at lights-out keeps a share of those receptors blocked. In a controlled trial, Drake and colleagues (2013) gave 400 mg six hours before bed: participants lost over an hour of measured sleep — and mostly did not report noticing.

Here is what the curve cannot tell you: whether you will feel it. Falling asleep quickly after an evening coffee proves little. Studies consistently find caffeine reduces deep, slow-wave sleep even when total sleep time looks normal. You can be unconscious for eight hours and still wake with the shallow version.

Your half-life is not average

Five hours is an average, and averages are where this chart stops being about you. Measured half-lives run from roughly 1.5 to 9.5 hours. Smoking speeds clearance sharply; oral contraceptives roughly double the half-life; pregnancy stretches it much further. The enzyme doing the work, CYP1A2, varies by genetics — fast and slow metabolisers drinking the same coffee are running different curves.

So the drawn curve is a template, not a verdict. A slow metaboliser's 15:00 coffee behaves like a fast metaboliser's 18:00 one. The x-axis is honest; the half-life powering it is borrowed. The only version of this chart that matters is the one your own mornings draw.

Reading your own curve

You cannot measure blood caffeine at home, but you can measure its shadow. Overnight HRV and resting heart rate respond to stimulants, and a morning recovery score aggregates them. Log your last caffeine of the day for two weeks — time and rough dose — then split the mornings into early-cut-off and late-cut-off days. One bad morning is noise, and alcohol, illness and late meals move the same numbers — so trust the pattern, not any single night. A consistent gap between the two piles is your half-life, expressed in the only currency that matters.