When does caffeine actually clear your system?
That afternoon coffee feels harmless hours later. But caffeine has a half-life that most people badly underestimate, and how fast it decays depends on factors that can shift your personal timeline by hours. Age, smoking, hormonal contraceptives, and pregnancy all alter the enzyme that clears caffeine from your blood. Enter your dose and a few personal factors to find out when it actually leaves.
Calculating clearance timeline…
Is the caffeine costing you sleep?
Calculate your weekly sleep debt and recovery cost.
How long does caffeine stay in your system?
Caffeine is cleared from the body through hepatic metabolism, primarily by the cytochrome P450 enzyme CYP1A2, which converts caffeine into paraxanthine, theobromine, and theophylline. Unlike alcohol, which follows zero-order elimination at a roughly fixed rate regardless of concentration, caffeine follows first-order kinetics: the rate of elimination is proportional to the amount present. This produces the characteristic exponential decay curve. The practical consequence is that caffeine never fully drops to zero in a single step; it falls by a fixed proportion each hour.
The half-life is the time required for blood caffeine concentration to fall by half. In healthy non-smoking adults aged 18 to 40 with no hormonal modifiers, the population mean is approximately 4 hours (range 1.5 to 9.5 hours). At a 4-hour half-life, a 200 mg dose (roughly two standard cups of filter coffee) still carries approximately 50 mg after 8 hours, 25 mg after 12 hours, and does not fall below 2 mg until around 24 hours. Sleep disruption research suggests that even 12 mg of residual caffeine can delay sleep onset in sensitive individuals. For a slow metaboliser with a 7-hour half-life, the same 200 mg dose takes closer to 42 hours to clear entirely.
The distinction between when you stop feeling alert and when caffeine has actually cleared your system is clinically important. Caffeine's subjective stimulant effects diminish as receptor tolerance shifts, but the molecule continues to occupy adenosine receptors and disrupt sleep architecture well after the subjective alertness has faded. Our caffeine consumption calculator puts your daily intake in population context, while this calculator addresses the specific question of elimination timing.
Why does caffeine half-life vary so much?
The primary driver of half-life variation is genetic. CYP1A2 is responsible for approximately 95% of caffeine metabolism in humans, and common polymorphisms in the CYP1A2 gene produce meaningfully different enzyme activity levels across the population. Fast metabolisers (CYP1A2*1A homozygotes) clear caffeine roughly twice as quickly as slow metabolisers (CYP1A2*1F C/C genotype), producing half-lives as short as 2.5 hours in the former and as long as 7 to 9 hours in the latter. Approximately 40 to 50% of adults in European populations carry the fast allele, though this calculator cannot test for genotype and applies population-average adjustments by group instead.
Beyond genetics, several environmental and physiological factors reliably shift CYP1A2 activity in predictable directions:
Smoking and vaping. Polycyclic aromatic hydrocarbons in tobacco smoke are potent inducers of CYP1A2 enzyme production. Smokers clear caffeine 40 to 50% faster than matched non-smokers, reducing typical half-life from around 4 hours to approximately 2.5 hours. This effect reverses within weeks of cessation, which is why recent ex-smokers sometimes report that caffeine feels noticeably stronger after quitting.
Oral contraceptives. Combined oral contraceptives (oestrogen plus progestin) competitively inhibit CYP1A2. A controlled trial by Abernethy and Todd (1985) demonstrated that OC use approximately doubled caffeine half-life, from a mean of 4 hours to approximately 7 to 10 hours depending on oestrogen dose. Progestin-only pills have a much smaller effect. This means OC users may experience stronger and longer-lasting caffeine effects from the same dose as a non-user.
Pregnancy. Progesterone and its metabolites progressively inhibit CYP1A2 across the course of pregnancy. In the first trimester, the effect is modest (half-life roughly 4.5 to 5 hours). By the third trimester, half-life extends to 10 to 18 hours, a factor-of-four increase from the pre-pregnancy baseline. This is why caffeine guidelines for pregnancy are considerably more conservative than for the general population: the same dose has a far greater and longer-lasting physiological effect.
Age. Hepatic clearance capacity declines gradually with age. Adults over 60 typically show half-lives 1.5 to 2 hours longer than young adults, reflecting reduced CYP1A2 activity and lower liver blood flow. The effect is gradual rather than a step change at any particular age.
The 3pm coffee problem
The most practically relevant implication of caffeine pharmacokinetics is what researchers sometimes call the afternoon caffeine effect. Consider a standard cup of filter coffee (approximately 95 mg) consumed at 3pm. By 10pm, seven hours have elapsed. The residual caffeine at bedtime depends entirely on the consumer's half-life:
| Population group | Half-life | Caffeine remaining at 10pm | Sleep disruption risk |
|---|---|---|---|
| Standard adult 18-40 | 4 hours | ~28 mg (29%) | Moderate (above 12 mg threshold) |
| Adults 60+ | 6 hours | ~42 mg (44%) | High |
| Smokers | 2.5 hours | ~14 mg (15%) | Low-moderate (just above threshold) |
| Combined OC users | 7 hours | ~48 mg (50%) | High |
| Pregnant, 3rd trimester | 15 hours | ~69 mg (73%) | Very high |
The threshold of 12 mg residual caffeine as a sleep-disrupting dose comes from controlled sleep studies reviewed in Grzegorzewski et al. 2022. This threshold is reached at approximately 5.2 hours for a standard adult (meaning a 3pm coffee crosses the disruption threshold around 8pm for most people, not at midnight). Smokers are the notable exception: their faster clearance means the same 3pm coffee may fall near or below the 12 mg threshold by 10pm. This is the pharmacological basis of the counterintuitive observation that smokers often report that caffeine does not affect their sleep as strongly as non-smokers report.
The practical cutoff for a sleep-undisrupted bedtime of 10pm, based on the 12 mg threshold, works backward as follows: for a standard adult (4h half-life), the last coffee should be consumed around 4.7 hours before bed, or roughly 5pm at the latest. For OC users (7h half-life), the cutoff is closer to 2pm. For smokers (2.5h half-life), even a 7pm coffee clears below the threshold by 10pm for most doses.
Frequently asked questions
Caffeine is considered fully cleared when less than 1% of the original dose remains in circulation. This takes approximately 6 half-lives. For a standard adult with a 4-hour half-life, full clearance from a typical dose occurs around 24 hours after consumption. For a slow metaboliser (7-hour half-life, common in oral contraceptive users), the same dose takes around 42 hours to fully clear. In third-trimester pregnancy, a 15-hour half-life means a morning coffee could still carry pharmacologically relevant levels the following morning. The half-life concept is the key: caffeine does not have a single "exit time." It tapers exponentially, with meaningful residual activity present long after subjective alertness has faded. (Source: Grzegorzewski et al. 2022, Frontiers in Pharmacology)
Caffeine half-life is the time required for your body to eliminate half of a given caffeine dose from your bloodstream. After one half-life, 50% remains. After two half-lives, 25% remains. After three, 12.5%, and so on. It is not a measure of when you stop feeling alert: caffeine's subjective stimulant effect diminishes partly through receptor adaptation independent of blood concentration, so you may feel back to normal while significant caffeine is still circulating. The half-life is the pharmacokinetic property that determines when the drug actually leaves your system, and it varies considerably across individuals and circumstances, which is why this calculator collects several personal factors before producing an estimate. (Source: Institute of Medicine, 2001; Statland and Demas, 1980)
Several factors can produce genuine resistance to caffeine's sleep-disrupting effects. The most common is the fast CYP1A2 metaboliser genotype, which produces a half-life of 2.5 to 3.5 hours rather than the population average of 4 hours. A fast metaboliser who drinks coffee at 3pm may have cleared enough caffeine by bedtime that sleep architecture is minimally disrupted. Smokers and vapers also clear caffeine significantly faster than non-smokers, putting them in a pharmacologically similar position. A third factor is long-term adenosine receptor upregulation in habitual caffeine users: the brain compensates for chronic receptor blockade by producing more adenosine receptors, which can reduce the stimulant effect of a given dose at the cost of more severe withdrawal headaches when the habit is interrupted. The person who says they can drink coffee at midnight and sleep fine is usually a fast metaboliser, a smoker, a very heavy habitual user, or some combination. (Source: Cornelis et al. 2006, JAMA; Astill et al. 2014)
Yes, and the effect is substantial. A controlled pharmacokinetic trial by Abernethy and Todd (1985) compared caffeine clearance in 18 women before and during use of combined oral contraceptives. They found that OC use reduced caffeine clearance by approximately 40%, extending half-life from a mean of roughly 4 hours to approximately 7 hours. In practical terms, this means that a woman on combined OC who drinks a large coffee at 3pm has roughly 50% of that caffeine still active at 10pm, compared to roughly 29% for a matched non-user. The mechanism is competitive inhibition of CYP1A2 by oestrogen metabolites. Progestin-only pills (the mini-pill) have a minimal effect on caffeine clearance. The interaction is dose-dependent: higher-oestrogen formulations produce a larger half-life extension. (Source: Abernethy and Todd, European Journal of Clinical Pharmacology, 1985)
Pregnancy produces a progressive and dramatic reduction in caffeine clearance, driven by progesterone-mediated inhibition of CYP1A2. In the first trimester, the effect is modest: half-life increases by roughly 0.5 hours from the pre-pregnancy baseline. In the second trimester, half-life typically extends to 7 to 10 hours. In the third trimester, caffeine half-life can reach 10 to 18 hours, with a population mean of approximately 15 hours. This represents a three to four times slower clearance rate than in non-pregnant adults. Caffeine crosses the placental barrier, and the foetus has virtually no CYP1A2 activity in early development, meaning it cannot metabolise caffeine independently. These pharmacokinetic facts underlie the conservative caffeine guidelines issued for pregnancy (typically 200 mg per day maximum in most national guidelines). Caffeine clearance returns to pre-pregnancy baseline within 2 to 3 weeks postpartum. (Source: Grzegorzewski et al. 2022; Institute of Medicine, 2001)
The answer depends on your personal half-life. Sleep disruption research suggests that residual caffeine above approximately 12 mg can delay sleep onset in sensitive individuals. Working backwards from a 10pm bedtime: for a standard adult with a 4-hour half-life, the last coffee should be consumed no later than around 5pm (4.7 hours before bed) for a 95 mg dose to fall below 12 mg by bedtime. For OC users (7-hour half-life), that cutoff is closer to 2pm for the same dose. For smokers (2.5-hour half-life), even a 7pm coffee may fall below the disruption threshold by bedtime. The calculator on this page computes your personalised cutoff based on your dose and half-life factors. Note that these thresholds are population averages: individual adenosine receptor sensitivity varies, so some people are meaningfully more sensitive or more resistant than these numbers suggest. (Source: Grzegorzewski et al. 2022; Drake et al. 2013, Journal of Clinical Sleep Medicine)
Yes, significantly. Polycyclic aromatic hydrocarbons in cigarette smoke are potent inducers of the CYP1A2 enzyme. Regular smokers typically have CYP1A2 activity 40 to 50% higher than matched non-smokers, producing a caffeine half-life of approximately 2.5 hours compared to the non-smoker mean of 4 hours. This is large enough to materially change the bedtime impact of an afternoon coffee: a smoker who drinks coffee at 3pm has roughly 15% of the dose remaining at 10pm, compared to about 29% for a non-smoker with the same baseline genetics. The enzyme induction effect disappears within weeks of smoking cessation, which is why many recent ex-smokers notice that caffeine feels stronger than it did when they were smoking. The effect applies to tobacco cigarettes and pipes, and appears to apply to vaping with nicotine, though the data on vaping specifically is more limited. (Source: Balogh et al. 1992; Grzegorzewski et al. 2022)
- Grzegorzewski J, Bartsch F, Herber K, Konig R. (2022). Pharmacokinetics of caffeine: A systematic analysis of reported data for application in metabolic phenotyping and liver function testing. Frontiers in Pharmacology, 12. DOI: 10.3389/fphar.2021.752826. [Meta-analysis of 141 studies, 4,714 individuals]
- Statland BE, Demas TJ. (1980). Serum caffeine half-lives: healthy subjects vs. patients having alcoholic hepatic disease. American Journal of Clinical Pathology, 73(3).
- 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).
- Institute of Medicine. (2001). Caffeine for the sustainment of mental task performance: formulations for military operations. National Academies Press.
- 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).
- This calculator provides pharmacokinetic estimates based on published population-average data. Individual variation is substantial, particularly due to CYP1A2 genetic polymorphisms which this tool cannot assess. Do not use this calculator to make clinical or medical decisions. If you are pregnant or trying to conceive, consult your healthcare provider about caffeine intake.