Caffeine half-life is genetic, not universal. Find the latest you can have your dose and still hit a clean sleep window — by your CYP1A2 phenotype, what else you take, and what bedtime you actually want.
Free. No login. Live math.
The popular advice that caffeine should be cut off ~8 hours before bed assumes an ~5 h half-life. About 10% of the population are slow metabolizers (CYP1A2 *1F variant) with half-lives closer to 7 h — for them, the cutoff is closer to 11 h. For fast metabolizers, it can be as short as 5 h. Genotype matters.
You can be a fast metabolizer and still be subjectively sensitive — adenosine-receptor variants (ADORA2A) drive perceived effect even when CYP1A2 clears the molecule quickly. The cutoff math models clearance, not subjective wakefulness. More on sleep-supplement interactions.
Drake 2013 in JCSM showed that 400 mg of caffeine 6 hours before bed reduced total sleep time by >1 hour. The 50 mg circulating threshold is a conservative estimate of the dose at which sleep architecture starts measurably degrading — but if you're caffeine-naive or sleep-fragile, even less can matter. Trim accordingly.
This uses first-order pharmacokinetics with a population-mean absorption assumption (absorption is rapid, tmax ≈ 30-45 min after ingestion). Fed-state absorption is slower than fasted. Modifiers are approximations from the published CYP1A2 literature, not personal pharmacogenomic data. Use this as a directional planning tool, not a clinical guarantee.
This tool is editorial. Wellness Radar takes no payment from supplement, coffee, or pharmacogenetic brands to promote any product. Where we publish sponsored content, it is clearly labeled.
It comes down to the CYP1A2 gene variant. About 50% of people are 'rapid' metabolizers (caffeine half-life ~4–5 hours), ~40% are 'intermediate' (~5–6 hours), and ~10% are 'slow' (~7–9+ hours). Genetic testing (23andMe, AncestryDNA + Promethease) reports this directly. Behavioral signal: if a 3pm coffee disrupts your sleep, you're likely slow. If you can drink coffee at 8pm and sleep fine, you're likely fast.
Plasma half-life averages 5 hours but ranges from 1.5 to 9.5 hours depending on genetics, pregnancy, oral contraceptives, smoking, and liver function. To clear caffeine to <10% of peak (often the threshold below which sleep architecture isn't measurably affected), allow 4–5 half-lives. For a 5-hour half-life, that's 20–25 hours from your last cup.
Caffeine blocks adenosine receptors. You can feel alert because caffeine masks adenosine buildup, but the moment caffeine clears, the accumulated adenosine hits you — and the drug-induced delay shifts your sleep onset. A 2013 study (Drake et al., J Clin Sleep Med) found 400mg caffeine 6 hours before bed reduced total sleep time by 1+ hours, even when subjects reported feeling fine.
Yes — typically 2–15 mg per 8 oz cup (vs 80–120 mg for regular). For most people that's negligible. For very slow metabolizers or people sensitive to caffeine for anxiety reasons, decaf within 6 hours of bedtime can still be enough to affect sleep latency.
No. Tolerance develops to the subjective stimulant effect (you don't feel as wired), but the physiological adenosine-receptor blockade and sleep disruption persist. Habitual coffee drinkers still show suppressed deep sleep when consuming caffeine in the afternoon — they just don't notice it consciously.
A typical pre-workout has 150–300 mg caffeine — equivalent to 2–3 cups of coffee. Taking that at 6pm with a 10pm bedtime gives you only 4 hours of clearance — about one half-life for an average metabolizer. You'll still have ~50% of peak caffeine in your system at bedtime. For evening training, switch to caffeine-free pre-workout (citrulline, beta-alanine, creatine work without caffeine) or train earlier.
Caffeine is caffeine — pharmacokinetics don't care about the source. 100 mg from matcha clears at the same rate as 100 mg from coffee. The L-theanine in tea blunts the stimulant edge subjectively, but doesn't change clearance or sleep impact at the same dose.