Caffeine Curfew Calculator

How to read this

The half-life is the whole game. Caffeine does not leave in one go, it halves. A 200mg pre-workout at 4pm is still 100mg at 9pm and 50mg at 2am. This page uses five hours, which is the population average, but people genuinely differ: the range runs from about four hours to about nine. If coffee has never once kept you awake you are probably at the fast end and these numbers are pessimistic. If an afternoon flat white reliably costs you sleep, you are at the slow end and the real figures are worse than what you see here.

Why there is no weight box. Caffeine is cleared by an enzyme in the liver, not by body mass, so asking your weight would look scientific and tell you nothing. A 60kg athlete and a 95kg one with the same genetics clear it at much the same rate.

Where the 50mg line comes from. Landolt and colleagues measured the sleep EEG directly and found that caffeine suppresses low-frequency delta activity, the slow-wave activity that makes deep sleep restorative, with effects persisting through the night from residual levels as low as 3.5 µmol/L. Drake and colleagues found that 400mg taken six hours before bed cut objectively measured total sleep time by more than an hour. Sleep researchers generally aim for no more than 30 to 50mg still on board at bedtime, which is where this page draws its line. The 20mg figure below it is the point where the effect becomes hard to detect at all.

You will probably not feel it. This is the part worth sitting with. The disruption shows up on an EEG whether or not you notice it, so falling asleep easily after an evening coffee is not evidence that it did no harm. Plenty of people who insist caffeine does not affect them are losing deep sleep every night, and the caffeine keeps working through the night rather than stopping once you are under.

Landolt HP, Dijk DJ, Gaus SE, Borbély AA (1995). Caffeine reduces low-frequency delta activity in the human sleep EEG. Neuropsychopharmacology 12(3), 229 to 238.
Drake C, Roehrs T, Shambroom J, Roth T (2013). Caffeine effects on sleep taken 0, 3, or 6 hours before going to bed. Journal of Clinical Sleep Medicine 9(11), 1195 to 1200.

Where you buy it matters more than how much you drink. Which? tested the high street in August 2026 and found a medium Costa cappuccino at 325mg, against 89mg for the same drink at Starbucks. That one cup is roughly four cans of Red Bull. Greggs came in at 235mg and Caffè Nero at 221mg, with Caffè Nero having roughly doubled since the 2023 study. The reason is the bean: Costa and Greggs blend robusta, which carries about twice the caffeine of the arabica used by Pret and Starbucks. Matcha is not a safe harbour either, ranging from 28mg at Pret to 99mg at Costa.

Athletes, watch your gels. A caffeinated gel late in a long Sunday session is a common and completely invisible reason for a bad night afterwards. A Maurten CAF 100 at 3pm behaves exactly like a strong coffee at 3pm.

High street figures from the Which? study published 25 August 2026. Everything else is a typical value and will vary with brew strength, bean and how heavy your hand is.

Why Combat Fuel are at the top of the list. They are the supplements Tom actually uses and has done for years, and TC Coaching has a commercial arrangement with them. That is why they head the supplement group rather than sitting in alphabetical order, so you should know it. Their caffeine figures are worked out exactly the same way as every other product here, and in their case that meant listing both pre-workouts lower than the number on the tub: the headline counts di-caffeine malate at its full weight and, on Dynamite, 200mg of TeaCrine, which is a stimulant but is not caffeine and does not behave like it.

Averages, not diagnosis. Individual clearance varies a lot, and pregnancy, the combined pill and some medications slow it considerably. If sleep is a real problem for you, that is a conversation with a GP, not a calculator.