How much caffeine is in your drink — and how long it lasts

Caffeine, measured — not guessed

Most caffeine pages stop at a single number. TrackCaffeine goes further: for every one of our 87 drinks we publish the exact milligrams, the strength per 100 ml, the equivalent in cups of coffee, and — uniquely — how much caffeine is left in your body after 1, 3, 6 and 12 hours. Those decay figures come from the same first-order pharmacokinetic model used in our CoffeeLog app, where caffeine halves roughly every 5.7 hours.

That second layer matters because the dose on the label is not the dose in your blood at bedtime. A 160 mg energy drink still leaves about 77 mg working six hours later. A large brewed coffee at 2 pm can keep more than 100 mg circulating near midnight. Our half-life calculator and last-cup-before-bed tool turn that into a clear answer: how much is in you now, and the latest moment you can have one more cup without wrecking your sleep.

Everything is sourced from public manufacturer data and USDA FoodData Central, published openly as a downloadable dataset, and labelled as estimates rather than medical figures. Start with a drink, run the numbers, then let CoffeeLog track it automatically.

Why does the milligram figure printed on a can answer the wrong question?

The milligram number on a can or label tells you how much caffeine is in that container, not how much is active in your body over time; that single figure does not account for elimination after you drink it. The label is a dose, not the circulating amount at bedtime or at any later moment.

The term "dose" is the amount of caffeine in a serving, and a can label reports the dose. The label is useful for inventory — for example Monster Energy lists 160 mg per 473 ml on the label — but it does not tell you how much remains in your bloodstream six or twelve hours later.

That difference matters because the effect you feel and the sleep impact depend on remaining caffeine, which falls over time; our pages compute remaining caffeine from the published dose so you can compare a 160 mg Monster against a 200 mg Cold brew in the same units. For example, Monster Energy (160 mg) and Cold brew (200 mg) are listed in our dataset and used in our decay examples to show remaining milligrams after fixed intervals.

Manufacturer numbers come from public packaging or company disclosures and are cross-checked against USDA FoodData Central when available, and the dataset is published under CC BY 4.0 so anyone can inspect the source of each milligram value (USDA FoodData Central, CC BY 4.0).

What does first-order elimination actually mean in practice for a normal day?

First-order elimination means the fraction of caffeine removed per hour depends on how much is present, so the more you start with the faster the absolute milligrams fall early on. First-order elimination is a pharmacokinetic process where the elimination rate is proportional to the current amount of substance.

Half-life is the time it takes for the amount to fall to half, and on this site we use a population-average half-life of 5.7 hours (342 minutes) in our model. Half-life is the time required for the concentration of a substance to decrease by 50% under first-order elimination.

The mathematical model we use is remaining = dose × 0.5^(hours ÷ 5.7), which gives a reproducible remaining-mg number for any drink and any hour value. For a concrete example, a 160 mg Monster Energy after six hours is: remaining = 160 × 0.5^(6 ÷ 5.7) = 160 × 0.5^(1.0526) ≈ 160 × 0.482 ≈ 77.1 mg, and that arithmetic is shown on each drink page and in our half-life calculator.

Another example: Cold brew (200 mg) after six hours is remaining = 200 × 0.5^(6 ÷ 5.7) = 200 × 0.482 ≈ 96.4 mg, which shows why high-dose drinks can still leave substantial caffeine many hours later. Our last-cup-before-bed tool uses the same equation to convert dose into a recommended time margin before a target bedtime.

The model is intentionally simple: it is first-order with a single half-life for population-average guidance, not a multifaceted clinical model; that makes it easy to compare drinks and to plug into tools like our side-by-side comparison and safe daily limit calculators.

Why does the same coffee affect two people differently?

People metabolise caffeine at different speeds because of genetics, lifestyle and physiology, so the same 126 mg Americano can leave very different amounts in different people after the same interval. CYP1A2 is a liver enzyme that metabolises caffeine, and genetic variants in the CYP1A2 gene change how quickly that process works.

CYP1A2 is a cytochrome P450 enzyme important for caffeine clearance, and factors that induce or inhibit it alter apparent half-life. Smoking induces CYP1A2 and tends to speed clearance, while pregnancy and many oral contraceptives inhibit CYP1A2 and can slow clearance, meaning identical doses produce different circulating mg over time.

To make that concrete: Espresso (double) is 126 mg; in a faster metaboliser the effective half-life might be shorter than 5.7 hours and the remaining mg after four hours could be much lower than in a slower metaboliser on oral contraceptives, who would retain more caffeine at the same time point. This is why our pages label numbers as estimates and why personal tracking (we mention CoffeeLog once) helps replace a population-average with a personalised pattern.

Non-genetic factors matter too: heavy smoking, certain medications and pregnancy status are documented modifiers of caffeine pharmacokinetics, and reviewers should consult clinical references and our guides such as the pregnancy and safety guides for context (Caffeine During Pregnancy).

What is "100 mg at bedtime" and what is it not?

"100 mg at bedtime" is a practical rule of thumb some people use to estimate low residual caffeine, not a clinical threshold or a guaranteed safe amount. It is a rule of thumb, not a clinical threshold.

Different organisations give safe intake guidance that is independent of bedtime thresholds: for example the U.S. FDA cites 400 mg/day for healthy adults as the reference daily upper bound (FDA, 400 mg/day). The research literature also shows timing matters: Drake et al. 2013 reported that 400 mg taken six hours before bedtime cost over an hour of sleep in a controlled study (Drake et al., 2013).

A pragmatic interpretation is: under our population-average model 100 mg remaining at bedtime is a non-zero dose likely to affect sleep for some people, but it is not a medical cut-off; use the last-cup-before-bed tool to convert doses such as Starbucks Pike Place Brewed (310 mg) or Latte (68 mg) into a timing estimate for your bedtime. For example, 310 mg at 10 hours later gives remaining = 310 × 0.5^(10 ÷ 5.7) ≈ 310 × 0.296 ≈ 91.8 mg, which illustrates how a large morning cup can still leave near-100 mg at night.

The caveat is that individual sensitivity varies: some people notice sleep effects at much lower residual doses while others are less sensitive; our tools aim to turn population averages into personal decisions, not to supply clinical judgement.

How is this site's dataset built and where can it be wrong?

The dataset is built from public manufacturer disclosures and USDA FoodData Central, compiled into an open CSV and published under CC BY 4.0 so every value is traceable. The dataset is a public compilation of manufacturer data and USDA FoodData Central entries, published as an open download.

Every drink page names the source for its milligram value and links back to the downloadable dataset so you can see the original citation; this transparency is designed to let you verify entries like Cold brew (200 mg), Bang (300 mg), Starbucks Caffè Americano (Grande) (225 mg) and Espresso (single) (63 mg). Users can report corrections via our corrections and contact page if a manufacturer updates a label.

The most common sources of error are portion-size mismatch, regional recipe differences and undocumented concentrate or extraction variations; brewed methods vary so Pour over (145 mg) and Drip coffee (96 mg) are method-sensitive examples from our table. When a shop lists a size but uses extra shots or a different brew ratio, the published value can under- or over-estimate true intake; that is why we treat values as estimates and provide per-100 ml strength where possible.

We publish the licence and source list so researchers and users can re-use and audit the dataset (USDA FoodData Central, CC BY 4.0), and you can explore drinks in the index such as Cold brew, Monster Energy, Bang, Starbucks Pike Place Brewed (Grande) and many more via the drink index. If you want to run numbers yourself, use our half-life calculator, last-cup-before-bed, safe daily limit and comparison tools to convert doses into remaining mg and timing guidance.