How Long Does Caffeine Last? Half-Life Calculator
Enter how much caffeine you took and how long ago, and this page tells you how many milligrams are still in your body. It reports the answer as an interval as well as a midpoint, because the published half-life is itself a range that varies several-fold between people. If you drink coffee every day, the same number can feel very different, because tolerance changes how adenosine receptors respond long after the plasma concentration has fallen.
Results appear here as you type.
Worked examples
These are produced by the same function your browser just ran, computed when this page was built.
| Case | Output |
|---|---|
| One 200 mg dose, healthy adult, 8 hours | Half-life applied (hours): 5; Published half-life interval: 3-7 h; Caffeine remaining (mg): 66; Plausible remaining range: 31.5-90.6 mg; Hours until under 50 mg: 10; Hours until under 100 mg: already below; Hours until under 150 mg: already below |
| One 200 mg dose, oral contraceptive, 8 hours | Half-life applied (hours): 10; Published half-life interval: 6-14 h; Caffeine remaining (mg): 114.9; Plausible remaining range: 79.4-134.6 mg; Hours until under 50 mg: 20; Hours until under 100 mg: 10; Hours until under 150 mg: already below |
| 400 mg afternoon dose, smoker, 6 hours | Half-life applied (hours): 2.5; Published half-life interval: 1.5-3.5 h; Caffeine remaining (mg): 75.8; Plausible remaining range: 25-121.9 mg; Hours until under 50 mg: 7.5; Hours until under 100 mg: already below; Hours until under 150 mg: already below |
How it is calculated
C(t) = D * exp(-ln(2) * t / t_half)
First-order exponential elimination. Time-to-threshold inverts the same expression. Every profile is reported as an interval as well as a midpoint, because the published half-life is itself a range.
Formula source: Standard one-compartment first-order elimination; see Temple JL et al., Front Psychiatry 2017;8:80 for the population half-life values used above.
Reference data on this page
Sources: Temple JL, Bernard C, Lipshultz SE, et al. "The Safety of Ingested Caffeine: A Comprehensive Review", Front Psychiatry 2017;8:80, doi 10.3389/fpsyt.2017.00080, PMID 28603504 — states an adult half-life of 3-7 h, that cigarette smoking doubles the clearance rate, that oral contraceptives tend to double the half-life, and that pregnancy lengthens it by about 8.3 h on average. Guidance context only: EFSA NDA Panel, "Scientific Opinion on the safety of caffeine", EFSA Journal 2015;13(5):4102.
https://pmc.ncbi.nlm.nih.gov/articles/PMC5445139/ · retrieved 2026-09-20 · citation checked
Limits of this tool
Estimates only. Every profile is a midpoint of a published interval, not a measurement of you. Genetic CYP1A2 variation, liver function, antibiotics, age and body mass shift half-life substantially, and some people sit far outside every interval listed here. The model covers a single oral dose: it does not track accumulation across repeated doses, tolerance to subjective effects, or the fact that adenosine-blocking effects persist after plasma levels fall.
How to use these numbers
The calculator returns three things: how much caffeine remains at a time you choose, how long until it falls below each of the planning thresholds, and the half-life applied. The thresholds are not medical cut-offs. They are the values most often quoted when people try to protect sleep, and you should treat them as starting points to test against your own experience rather than as a rule.
The most useful reading is the reverse one: pick your intended bedtime, count backwards, and find the latest time your morning coffee stops mattering. Take 200 mg at 8:00 am. On the 5-hour midpoint the model uses for a healthy adult, about 29 mg is left by 10:00 pm. On the 7-hour end of the published interval, the same dose leaves about 50 mg, which is the threshold this page shows for sleep. Same dose, same clock time, a two-fold difference in what is still in your body.
Caffeine and sleep: why the clock matters more than the cup size
Caffeine does not just reduce total sleep time; it can shrink deep sleep and slow sleep onset even when people do not notice it. A review of clinical sleep literature notes that caffeine taken within a few hours of habitual bedtime tends to reduce slow-wave sleep, while habitual users often under-report that effect because alertness itself masks the underlying disruption.
The practical consequence is that the timing of the dose matters as much as the amount. A 300 mg dose at 4:00 pm can still leave more than 37 mg in a healthy adult at 11:00 pm, which is the sleep-planning threshold used on this page. For people who go to bed later, the same afternoon coffee may be acceptable; for people who sleep early, it may not. The calculator makes the trade-off visible by letting you move the time slider.
How much caffeine is really in your drink or snack
Beverage labels are often poor proxies for what you actually drink. A standard 8 oz cup of home-brewed coffee typically contains 70 to 140 mg of caffeine, but a 16 oz chain-store latte can exceed 300 mg once the larger serving size is included. A single espresso shot is commonly 50 to 75 mg, but some specialty roasts push a double shot well above 150 mg.
Tea is not automatically safer. A 240 ml cup of black tea usually carries 40 to 70 mg, but strong masala chai or matcha can reach 80 mg or more. Green tea is often closer to 30 to 50 mg. Energy drinks are the widest range: small cans may contain only 30 mg, while large functional cans or powder mixes can exceed 400 mg per serving.
Hidden sources are common. Dark chocolate averages 30 to 50 mg per 30 g serving, and some high-caffeine baking or confectionery products advertise 100 mg per piece. Over-the-counter migraine or headache pills often contain 65 mg per tablet, and pre-workout supplements may carry 150 to 400 mg per scoop. Many people omit these items when they log intake, so the effective dose is higher than the coffee count alone suggests. Even "decaffeinated" coffee is not a perfect zero: most brewed decaf cups still contain 2 to 15 mg, which is negligible for most calculations but can matter if you are very sensitive or using multiple cups late in the day.
Why clearance profiles differ: pregnancy, smoking, contraception, and genes
Almost all caffeine is cleared by one liver enzyme, CYP1A2, and that enzyme is easily changed by other things. Compounds in tobacco smoke induce it, which is why a comprehensive review reports that smoking roughly doubles the rate at which caffeine is cleared. Combined oral contraceptives tend to inhibit it, and the same review reports that they roughly double the half-life. In pregnancy the review reports an average increase of about 8.3 hours on top of the adult figure.
Liver disease slows clearance too, but the published numbers are spread too widely to put a single figure here, so that group has no profile and the calculator will not guess one. Age matters as well: adolescents and older adults often clear caffeine more slowly than younger adults, but the evidence is inconsistent enough that this page does not present a separate figure.
Genetics matters on top of all of this. People carrying less-active CYP1A2 variants clear caffeine slowly enough that a single espresso can produce hours of measurable jitteriness, while fast metabolisers report drinking coffee at night and sleeping normally. The calculator cannot see your genotype, so the honest approach is to run your own test: take a dose you know, and note when subjective effects end.
Where the reference data comes from
The four figures on this page are derived, not quoted directly. The source states an adult half-life of 3-7 hours; the midpoint, 5 hours, is the healthy-adult profile. The smoker profile divides that midpoint by two, the oral-contraceptive profile multiplies it by two, and the pregnancy profile adds the 8.3-hour average increase the source reports. Each profile keeps the published interval alongside it, and the results panel shows what that interval does to your answer.
This page deliberately avoids giving a "safe" milligram figure, because the widely quoted 400 mg daily ceiling is a population guidance value covering pregnancy and other conditions differently, and it is not something a half-life model can compute.
Caffeine timing for shift work and jet travel
Shift workers live in the gap between circadian rhythm and caffeine pharmacokinetics. A night-shift worker who drinks 200 mg at the start of a shift may still have measurable caffeine in the body by the time they try to sleep during daylight hours. The same 5-hour half-life that governs a morning coffee applies at midnight, which means the body does not reset the clock just because the shift schedule changed.
For people travelling across time zones, caffeine can be a short-term aid for staying awake during the destination daytime, but it will also delay sleep onset if taken too late in the local afternoon. A useful rule is to treat the local bedtime as the fixed point and apply the calculator backwards, the same way you would at home.
Medications and supplements that interact with caffeine clearance
Caffeine is not the only thing CYP1A2 processes. Several commonly prescribed drugs inhibit the same enzyme, which means they can slow caffeine clearance without changing the caffeine dose. Antibiotics such as ciprofloxacin and norfloxacin are well-documented inhibitors, as is the antiarrhythmic mexiletine. Some antidepressants, including fluvoxamine, also raise caffeine levels by reducing clearance.
Over-the-counter supplements can do the same thing. Grapefruit juice does not act on CYP1A2, but it does inhibit CYP3A4, which handles some other stimulants; for caffeine specifically, the interaction is small. More relevant are herbal preparations that claim to support liver detoxification: if they induce CYP enzymes, they may speed caffeine clearance modestly, while inhibitors in complementary medicines can slow it. The safest approach is to treat any new medication or supplement as a possible modifier until you see how the calculator values change for you.
Building tolerance: what the half-life model does not show
Half-life describes clearance from plasma, not how you feel after the plasma level has fallen. Regular drinkers develop tolerance to the alertness and jitteriness effects, which means the same dose produces a smaller subjective response over time. That tolerance does not remove the sleep-system interaction: even if you feel fine, enough adenosine blockade may remain to reduce deep sleep.
Caffeine dependence is defined more by withdrawal than by dose. People who consume caffeine every day often experience headache, fatigue, and reduced alertness if they skip their usual morning cup, and those symptoms can begin within 12 to 24 hours. From a planning standpoint, that means the calculator is most useful when you are deciding whether to skip or delay a dose, because the withdrawal window overlaps the clearance window.
How to plan across morning, afternoon, and evening doses
Most people do not drink caffeine once per day and never again. Espresso at breakfast, a cola at lunch, and an iced coffee in the afternoon create overlapping elimination curves, and the calculator on this page is built for a single dose. If you want to approximate a repeated-dose day, calculate each drink separately and add the remaining milligrams at the clock time you care about. That gives an upper bound rather than an exact answer, but it is usually conservative enough for sleep planning.
A common mistake is to treat all afternoon coffee as the same risk. A 100 mg drink at 2:00 pm leaves less than 25 mg by 10:00 pm on the 5-hour midpoint, but the same drink at 5:00 pm can leave more than 37 mg at the same bedtime. The later the dose, the more sensitive the answer becomes to the half-life interval. That is why the calculator reports an interval rather than a single number.
If you are trying to protect deep sleep rather than just falling asleep, aim for the sleep threshold to clear at least two hours before bedtime. The 50 mg sleep threshold on this page is a planning value, not a clinical cut-off, but keeping it below that level for a full two hours gives the adenosine system more time to recover its normal sensitivity before lights-out.
Questions
How much caffeine is in a cup of coffee?
An 8 oz (240 ml) brewed cup commonly contains 70-140 mg; espresso shots about 50-75 mg; a "mug" serving is often double an 8 oz cup.
Why does smoking change the result?
Compounds in tobacco smoke induce CYP1A2, the liver enzyme that clears most caffeine, so caffeine is cleared roughly twice as fast in smokers.
Does this predict sleep quality?
No. It estimates milligrams remaining. The sleep-impact threshold is a common planning value, not a clinical cut-off.
Does body weight change the half-life?
Half-life is a clearance time, not a concentration, so weight is not in the formula. A larger person may have more total caffeine per kilogram if they drink the same cup, but the clock until half remains is similar.
Can I speed up caffeine clearance?
Hydration and food slow absorption, not elimination. Once caffeine is in the bloodstream, clearance is set by liver enzyme activity, and there is no evidence that common remedies meaningfully shorten the half-life.
Why do I still feel alert after the calculator says caffeine should be gone?
Plasma concentration and subjective alertness are not identical. Adenosine receptor occupancy, tolerance, and withdrawal reversal can all make you feel wired even when the remaining milligram figure is small.
What about decaf coffee?
Most decaf still contains 2 to 15 mg per cup. It is unlikely to move the calculator much, but it can matter if you are very sensitive or use multiple cups to wind down at night.
Does tea affect sleep differently from coffee?
Tea contains less caffeine per serving and adds L-theanine, but the page does not model subjective calmness or sleep architecture. If you are using the calculator for timing, enter the estimated tea dose and the same half-life profile.