How we calculate
Our calculators estimate how much caffeine is still in your body at a given time. Here is the model, the numbers it uses, where they come from and what it cannot know about you.
The model: caffeine halves at a steady rate
Caffeine taken by mouth is absorbed quickly and almost completely (Blanchard & Sawers 1983). The body then clears it gradually. We model that clearance with a half-life: after one half-life, half the dose is left; after two, a quarter; and so on.
In numbers: caffeine left = dose × 0.5(hours ÷ half-life). For example, 100 mg taken 6 hours earlier leaves about 25 mg with a 3-hour half-life, 44 mg with 5 hours and 59 mg with 8 hours.
Why we show three half-lives, not one
Half-life differs a lot between people. The European Food Safety Authority gives an average of about 4 hours in healthy adults, with a range of about 2 to 8 hours. In one study of 16 healthy men it ranged from about 2.3 to 9.9 hours (Blanchard & Sawers 1983). Researchers in a 2025 sleep trial describe it as typically 3 to 6 hours (Gardiner et al. 2025).
So every result shows a range: a fast (3 h), average (5 h) and slow (8 h) clearance. These three points are our own modelling choice inside the published range. Our “average” of 5 hours is a little more cautious than EFSA's 4-hour average.
Things that make caffeine last much longer
Some situations push half-life well beyond the slow end of our range. In published studies:
- Pregnancy: about 8.3 hours on average (range 3 to 16) in one study, versus 3.4 hours in non-pregnant adults (Knutti et al. 1981).
- Long-term use of low-dose estrogen birth-control pills: about 7.9 versus 5.4 hours in one small study (Abernethy & Todd 1985).
- The antidepressant fluvoxamine: half-life rose from about 5 to 31 hours in one study (Jeppesen et al. 1996).
- Alcohol-related liver disease: 60 and 168 hours in two patients (Statland & Demas 1980).
- Smoking works the other way: heavy smokers clear caffeine faster, and clearance falls after they quit (Faber & Fuhr 2004).
If any of these apply to you, our numbers will underestimate what is left. Ask a doctor or pharmacist what is right for you.
Time to take effect
In a small study of young men given caffeine in a drink, blood levels peaked about 30 minutes later (Blanchard & Sawers 1983). MedlinePlus says caffeine reaches its peak in the blood within about an hour, and EFSA says the stimulating effects may begin 15 to 30 minutes after you take it. Coffee sipped slowly or with food can take longer. Our timing calculator suggests drinking about 45 minutes before you need to be alert; that is a rounded planning choice, not a measured value for you.
The 50 mg bedtime target
The cutoff calculator works backwards from bedtime to the latest time that leaves 50 mg or less. No study names a safe amount of caffeine at bedtime, so this is our assumption, not a published threshold. We chose it because in a 2025 trial 100 mg taken 4 hours before bed had no significant effect on measured sleep (Gardiner et al. 2025), and our model leaves a little more than 50 mg from that dose.
Sleep studies also show that some people sleep worse even when they don't notice it, so treat any cutoff as a starting point. See caffeine and sleep for what the studies found at different doses.
What the model leaves out
- Absorption: the “left at bedtime” numbers start clearing caffeine from the moment you drink it; the cutoff calculator only adds a fixed 45-minute allowance. Real absorption varies.
- Your personal half-life: we can't measure it, so we show a range.
- Several drinks add up: the timing calculator sums what each drink leaves at bedtime.
- How you feel: the same amount affects people differently, and tolerance changes how it feels.
Sources
- Blanchard J, Sawers SJ. (1983). The absolute bioavailability of caffeine in man. European Journal of Clinical Pharmacology, 24(1), 93–98. PMID 6832208 ↗
- Blanchard J, Sawers SJ. (1983). Comparative pharmacokinetics of caffeine in young and elderly men. Journal of Pharmacokinetics and Biopharmaceutics, 11(2), 109–126. PMID 6886969 ↗
- European Food Safety Authority (2015). EFSA explains risk assessment: Caffeine. European Food Safety Authority (doi:10.2805/618813). efsa.europa.eu (PDF) ↗
- Gardiner CL, Weakley J, Burke LM, et al. (2025). Dose and timing effects of caffeine on subsequent sleep: a randomized clinical crossover trial. Sleep, 48(4), zsae230. doi:10.1093/sleep/zsae230 ↗
- MedlinePlus (U.S. National Library of Medicine) (2021). Caffeine: MedlinePlus. MedlinePlus, U.S. National Library of Medicine. medlineplus.gov ↗
- Knutti R, Rothweiler H, Schlatter C. (1981). Effect of pregnancy on the pharmacokinetics of caffeine. European Journal of Clinical Pharmacology, 21(2), 121–126. PMID 7341280 ↗
- Abernethy DR, Todd EL. (1985). Impairment of caffeine clearance by chronic use of low-dose oestrogen-containing oral contraceptives. European Journal of Clinical Pharmacology, 28(4), 425–428. PMID 4029248 ↗
- Jeppesen U, Loft S, Poulsen HE, et al. (1996). A fluvoxamine-caffeine interaction study. Pharmacogenetics, 6(3), 213–222. PMID 8807660 ↗
- Statland BE, Demas TJ. (1980). Serum caffeine half-lives. Healthy subjects vs. patients having alcoholic hepatic disease. American Journal of Clinical Pathology, 73(3), 390–393. PMID 7361718 ↗
- Faber MS, Fuhr U. (2004). Time response of cytochrome P450 1A2 activity on cessation of heavy smoking. Clinical Pharmacology & Therapeutics, 76(2), 178–184. PMID 15289794 ↗