Updated: 2026-09-07 · 6 min read · First-order decay · EFSA and FDA sourced
Caffeine leaves your body on a half-life of about five hours — anywhere from 1.5 to 9.5 depending on you. Enter what you drank and when. The calculator shows how much is still circulating and what will be left at bedtime, against what the sleep trials actually measured rather than folklore.
STILL IN YOU NOW—
AT BEDTIME—
DOWN TO 25 MG AT—
SHARE OF DAILY 400 MG CAP—
Two things this calculator cannot know — and neither can any other
Your input is probably wrong by more than the maths. The same coffee, from the same
outlet, on six consecutive days, measured between 259 and 564 mg. Single espressos have been
measured up to 322 mg, and four of twenty commercial shots exceeded 200 mg. The menu item
is not a dose.
Known factors explain about 37% of the variation. Coffee intake, smoking, oral
contraceptives and sex together account for just over a third of the difference in caffeine
clearance between people. The rest is not explained by anything a form can ask you.
So read the output as a shape rather than a measurement. The shape is still worth having: it is
the difference between “my coffee wore off hours ago” and “a third of it is still circulating”.
What the sleep trials actually found
There is no published milligram threshold at which caffeine starts affecting sleep. Any
calculator telling you that you will be “under the safe level” by nine o'clock invented that level.
What does exist is dose-and-timing evidence, and it is less forgiving than the usual advice.
Dose
Hours before bed
Measured effect
100 mg
4 hours
No significant effect on objective or subjective sleep
400 mg
4 hours
50.6 min less sleep, 14.2 min longer to fall asleep, 9.5% lower efficiency
400 mg
8 hours
29.1 min more time awake in the night, 6.9% lower efficiency
400 mg
6 hours
Over an hour of sleep lost — and the sleepers did not notice
107 mg (one coffee)
8.8 hours
Cut-off derived from a 24-study meta-analysis
Gardiner and colleagues 2025 (23 men); Drake and colleagues 2013 (12 adults);
Gardiner and colleagues 2023, meta-analysis of 24 studies.
The finding worth changing your habits over
In the 2013 trial, 400 mg taken six hours before bed cut objectively measured sleep by
more than an hour — while the sleepers' own diaries did not register a significant change. They lost
the sleep and did not feel it. The authors named this missing subjective signal as a likely reason
people ignore caffeine advice: the damage is real and the evidence of it is invisible from the
inside.
If you are cutting caffeine out entirely
Withdrawal starts 12–24 hours after the last dose, peaks at 20–51
hours, and runs 2–9 days. The headache can persist for up to three weeks.
Around half of people get a withdrawal headache in controlled abstinence studies, and
about 13% are impaired enough for it to interfere with normal life. Population surveys put the
real-world figure far lower — the two numbers are measuring different things.
You do not need a big habit. Withdrawal has been produced from 100 mg a day, and
after as little as three days of use.
The ten validated symptoms: headache, fatigue, low energy, reduced alertness, drowsiness, reduced
contentedness, depressed mood, poor concentration, irritability, and feeling foggy.
Tapering is standard clinical advice, and a six-week reduction programme has randomised support
for sustaining lower intake. Whether tapering reduces withdrawal severity versus stopping
outright rests on one small pilot — so it is sensible rather than proven.
Half of it is gone after about five hours, three-quarters after ten, and roughly 94% after twenty. But the individual range is 1.5 to 9.5 hours: a heavy smoker clears a coffee in about three hours, while someone in late pregnancy takes three to four times longer than they normally would.
Is it true that caffeine lingers longer as you get older?
No. A direct comparison of men around 20 and around 71 found time to peak, peak concentration and bioavailability essentially identical, with clearance slightly higher in the older group. Half-life is not affected by adult age. This is one of the most repeated claims about caffeine with no support behind it.
Does a DNA test tell me if I am a slow caffeine metaboliser?
Not usefully. The variant those tests read changes how much your CYP1A2 enzyme can be induced, which is mainly visible in smokers and heavy coffee drinkers. In non-smokers, no difference in enzyme activity has been found between the genotypes, and a 786-person study found no relevant functional polymorphism once other factors were accounted for. This calculator deliberately does not ask.
Does caffeine dehydrate you?
Not at ordinary doses. A crossover trial in 50 habitual coffee drinkers using deuterium-oxide total body water found no significant difference between four cups of coffee and the same volume of water across total body water, urine volume, specific gravity or osmolality.
Is decaf caffeine-free?
No — 2 to 15 mg per cup, measured as high as 17.7 mg. Fine for most people, but not nothing if you are eliminating caffeine completely or are unusually sensitive.
How much caffeine is too much in a day?
Both the FDA and EFSA put 400 mg a day for healthy adults at the level not generally associated with negative effects, with EFSA adding a 200 mg limit for a single dose. EFSA advises 200 mg a day in pregnancy; Health Canada says 300 mg; the FDA declines to set a number and says to ask your provider.
General educational information from public health literature and peer-reviewed pharmacokinetics. Not medical advice. The model assumes first-order elimination, which is well supported up to about 400 mg per dose; above roughly 500 mg it will under-estimate what is left. If you are pregnant, have liver disease, or take fluvoxamine, ciprofloxacin or cimetidine, your clearance may differ far more than this tool allows for.