Caffeine content of every drink we track

How should I read a caffeine figure — per serving versus per 100 ml, and why are a small strong drink and a large weak one different problems?

The simplest answer: a per-serving caffeine figure tells you the total caffeine you will ingest when you finish that serving, and a per-100 ml figure tells you the concentration of caffeine in the beverage; both matter for different choices. Serving is the amount of drink listed for that product and concentration is the amount of caffeine per 100 millilitres; the serving number decides dose and the per-100 ml number decides intensity.

Compare two real examples from the dataset: a single Espresso (30 ml) contains 63 mg total caffeine as shown on its page Espresso (single), 63 mg, and Cold brew (240 ml) contains 200 mg total as shown on its page Cold brew, 200 mg. The espresso is tiny (30 ml) and concentrated (63 mg in 30 ml) and the cold brew is large (240 ml) and high total dose (200 mg in 240 ml), so a single espresso is a concentration problem while the cold brew is a dose problem; both of those sentences cite the dataset by linking each drink page.

When you see a per-100 ml figure you are seeing concentration, which is useful if you want to compare strength independent of cup size; Pour over is 145 mg per 240 ml and shows 60.4 mg/100 ml on its page Pour over, 145 mg, which is a concentration fact from the dataset. When you see a per-serving figure you are looking at dose, which is what matters to a daily limit or to sleep timing; Starbucks Pike Place Brewed (Grande) is 310 mg per 473 ml as listed on its page Starbucks Pike Place Brewed (Grande), 310 mg, and that 310 mg is the dose fact from the dataset and determines how much remains in your body hours later.

Caveat: two drinks can have the same per-100 ml but very different servings and thus very different total caffeine; Monster Ultra and Monster Energy each have similar concentration around 31.7–33.8 mg/100 ml as listed on their pages Monster Ultra, 150 mg and Monster Energy, 160 mg, but their servings and totals are the numbers you should add to your daily intake. This is a dataset fact because the per-100 ml entries come from the dataset pages cited in the same sentence.

What actually varies inside each category?

Different physical and procedural variables drive caffeine in each category; this answer lists the key variables for brewed coffee, espresso, coffee-shop menu drinks, tea, soda, energy drinks and pre-workout powders. Category is the grouping used in the dataset and the dataset groups 87 drinks into eight categories; the site index shows category counts and you can view the full index at the drink index.

What varies in brewed coffee?

Brewed coffee varies by bean, grind, brew time and water-to-coffee ratio, which together change both concentration and total serving dose. Drip coffee in the dataset is 96 mg per 240 ml on its page Drip coffee, 96 mg, and Pour over is 145 mg per 240 ml on its page Pour over, 145 mg, so the same 240 ml serving can differ by 49 mg because those brewing variables change extraction efficiency; the numbers and their pages are cited in the same sentence as dataset statistics.

Cold extraction methods concentrate differently: Cold brew is 200 mg per 240 ml as listed on its page Cold brew, 200 mg and Nitro cold brew is 215 mg per 240 ml as listed on its page Nitro cold brew, 215 mg, and those are dataset facts showing that long cold maceration extracts a high total dose even at room-temperature brewing.

What varies in espresso?

Espresso varies by dose (how much ground coffee is used), extraction (how long and how much water passes through) and basket or machine standard; espresso dose is not fixed across shops or machines. Espresso (double) is 126 mg in 60 ml on its page Espresso (double), 126 mg, and Flat white is 130 mg in 160 ml on its page Flat white, 130 mg, and these numbers show that shot count and dilution change the serving total even within the espresso category; both numbers are dataset facts cited in the sentence.

Ristretto and single espresso can show identical total caffeine with different volumes: Ristretto is 63 mg in 20 ml on its page Ristretto, 63 mg and Espresso (single) is 63 mg in 30 ml on its page Espresso (single), 63 mg, demonstrating that extraction style changes concentration without changing total dose according to the dataset entries cited here.

What varies in coffee-shop menu drinks?

Coffee-shop menu drinks mostly vary by shot count and by the chain’s standard serving size; customers often assume a drink name fixes caffeine, but our data show size and number of shots are the actual levers. Starbucks Caffè Americano (Grande) is 225 mg per 473 ml as listed on its page Starbucks Caffè Americano (Grande), 225 mg and Starbucks Caffè Latte (Grande) is 150 mg per 473 ml as listed on its page Starbucks Caffè Latte (Grande), 150 mg, which shows two drinks with the same serving volume can differ by 75 mg because the Americano has more concentrated espresso shots in the Grande size; both numbers are dataset facts with links to the dataset pages.

Chain variability example: Dunkin' Original Blend (Medium) is 210 mg per 414 ml on its page Dunkin' Original Blend (Medium), 210 mg and Dunkin' Latte (Medium) is 119 mg per 414 ml on its page Dunkin' Latte (Medium), 119 mg, showing that a milk-based latte at the same size can contain substantially less caffeine than plain brewed coffee at the same size; these are dataset-sourced numbers cited here.

What varies in tea?

Tea varies by leaf type, amount of leaf, steep time and whether the tea is whole-leaf, bagged or powdered; these choices change total caffeine in the serving. Matcha (1 tsp) is 70 mg per 240 ml as listed on its page Matcha (1 tsp), 70 mg and Green tea is 28 mg per 240 ml as listed on its page Green tea, 28 mg, demonstrating that powdered matcha concentrates both leaf and caffeine compared to infusion-style green tea; the dataset pages are the source cited for those numbers in the same sentence.

Tea type examples: Black tea is 47 mg per 240 ml on its page Black tea, 47 mg and Yerba mate is 85 mg per 240 ml on its page Yerba mate, 85 mg, and those dataset facts show that different traditional teas deliver different doses even at identical serving volumes.

What varies in soda?

Soda tends to be low concentration compared with coffee, and in some markets caffeinated sodas are capped by regulation which limits how much caffeine a can or bottle may contain. Coca-Cola (330 ml) contains 32 mg per 330 ml as listed on its page Coca-Cola (330 ml), 32 mg and Mountain Dew (355 ml) contains 54 mg per 355 ml as listed on its page Mountain Dew, 54 mg, both dataset values showing soda concentrations are generally small compared with coffee; the dataset pages are cited as sources.

Caveat: some sodas like Club-Mate are relatively concentrated: Club-Mate (330 ml) is 66 mg per 330 ml on its page Club-Mate, 66 mg, and that is a dataset fact showing exceptions exist within the soda category.

What varies in energy drinks?

Energy drinks vary by labelled caffeine content, the presence of other stimulant ingredients and by whether the maker lists a concentration per volume or a total per can; label accuracy matters because some products are high-dose. Bang (473 ml) is 300 mg per 473 ml on its page Bang, 300 mg and Red Bull (250 ml) is 80 mg per 250 ml on its page Red Bull (250 ml), 80 mg, both numbers from the dataset showing that energy drinks span a large range of total dose by serving in the dataset.

Label wording caveat: some energy drinks list proprietary blends or "energy blends" for other active ingredients while still giving caffeine as a number; Celsius is 200 mg per 355 ml as listed on its page Celsius, 200 mg, and that dataset value is the figure to use when tracking dose regardless of other blend wording.

What varies in pre-workout supplements?

Pre-workout powders vary by scoop size and by whether a product lists caffeine per scoop or per serving measured with water; scoop is the unit users must check to get the true dose. Pre-workout (1 scoop) is 200 mg per 300 ml on its page Pre-workout (1 scoop), 200 mg and C4 Original (1 scoop) is 150 mg per 300 ml on its page C4 Original (1 scoop), 150 mg, both dataset values, so swapping scoop sizes changes total caffeine consumed even when the volume is the same.

Caveat: pills and gum provide the caffeine dose directly as a number without liquid volume; a Caffeine pill (200 mg) is 200 mg on its page Caffeine pill (200 mg), 200 mg and Caffeine gum (1 piece) is 40 mg on its page Caffeine gum (1 piece), 40 mg, dataset facts showing these forms make arithmetic simpler because the serving is the dose itself.

Where do our numbers come from and how large is the honest error bar?

All drink caffeine numbers in this index come from our open dataset, which is published under a CC BY 4.0 licence and available at the dataset page; CC BY 4.0 is the licence on our dataset as noted at CC BY 4.0. The dataset is the source of every milligram and serving size in this index, and the dataset page the open dataset lists provenance and links for each entry.

Honest error bar means the combination of natural variability and measurement or labelling uncertainty; we do not invent numerical uncertainty in the index because the dataset lists point estimates for 87 drinks, but typical real-world brewed coffee variability can exceed tens of percent due to brewing variables explained earlier, and those statements are grounded in the multiple brewed coffee examples in the dataset such as Drip coffee (96 mg) Drip coffee, 96 mg and Pour over (145 mg) Pour over, 145 mg which show the scale of variation across similar servings in the dataset.

When we report a canned or labelled product we use the manufacturer or chain published figure where available and list that source on the individual drink page; Starbucks Cold Brew (Grande) is 205 mg per 473 ml on its page Starbucks Cold Brew (Grande), 205 mg and Dunkin' Cold Brew (Medium) is 260 mg per 710 ml on its page Dunkin' Cold Brew (Medium), 260 mg, and those are dataset values drawn from publicly posted chain nutrition data as documented on each drink page.

Caveat on accuracy: for manufactured products such as sodas and energy drinks, label accuracy is usually good but different batches and markets can vary; Red Bull (250 ml) is 80 mg per 250 ml on its page Red Bull (250 ml), 80 mg and Red Bull (355 ml) is 114 mg per 355 ml on its page Red Bull (355 ml), 114 mg, both dataset facts that show manufacturers publish values separately for different can sizes and that those published numbers are the ones we record.

Correction process: if you see an error on a drink page you may request an update through our corrections channels at corrections and contact, and the dataset provenance and licence are documented at method, sources and author and the dataset index the open dataset.

How do I use the decay model to turn any of these numbers into a bedtime decision?

Use first-order elimination with a population-average half-life of 5.7 hours: remaining = dose × 0.5^(hours ÷ 5.7). First-order elimination is the pharmacokinetic model we use and half-life is the time required for the body to remove half of the circulating caffeine; these are definitional sentences and the half-life used is the population average we apply as the site model.

Example arithmetic: if you drink a Latte (240 ml) with 68 mg as listed on its page Latte, 68 mg and you want to know how much caffeine remains after 6 hours, compute remaining = 68 mg × 0.5^(6 ÷ 5.7) = 68 mg × 0.5^(1.0526) = 68 mg × 0.484 ≈ 32.9 mg, and that arithmetic uses the dataset dose and the model half-life. This calculation shows the explicit mg arithmetic so you can see that 68 mg becomes approximately 32.9 mg after 6 hours using our model; the Latte page is cited as the dataset source for the initial dose.

Another example with a high-dose drink: if you finish a Bang (473 ml) containing 300 mg as listed on its page Bang, 300 mg four hours before bed, remaining = 300 mg × 0.5^(4 ÷ 5.7) = 300 mg × 0.5^(0.7018) = 300 mg × 0.615 ≈ 184.5 mg, which is the model arithmetic using the dataset dose and half-life. That residual 184.5 mg is the amount our model predicts will remain in circulation at sleep time and you can compare that to guidance such as the FDA 400 mg/day limit (which is cited below) when making a bedtime decision; the Bang page is cited as the dataset source for the starting dose.

Benchmarks and sources: the U.S. Food and Drug Administration notes 400 mg per day as a recommended upper limit for healthy adults and you can read that guidance at the FDA page FDA, 400 mg/day for healthy adults, which is the external policy number to reference when you total daily intake. Sleep timing evidence: Drake et al. 2013 found that 400 mg of caffeine given six hours before bedtime cost subjects more than an hour of sleep in a controlled study, and you can read that study at the Journal of Clinical Sleep Medicine page Drake et al. 2013; use that study when deciding how late a high total dose might meaningfully disturb sleep.

How to use the site tools: convert any drink to a sleep-ready decision with our calculators — use the half-life calculator at caffeine half-life calculator, estimate the last cup you can have before bed at latest cup before bedtime, compare total intake against weight-based guidance at safe daily limit by body weight, and put two drinks side-by-side at caffeine comparison. Each tool links back to dataset entries for the drinks you pick so the calculators use the exact mg values shown on the drink pages.

Practical rule-of-thumb using the model: if you want less than roughly 50 mg remaining at bedtime, then a 100 mg dose should be taken at least about 5.7 hours earlier because 100 mg × 0.5^(5.7 ÷ 5.7) = 100 mg × 0.5 = 50 mg; this example uses the half-life formula and the model half-life and is provided to illustrate how the arithmetic maps to practical timing with dataset doses and the model half-life.

Where can I compare entries and run the math for specific drinks?

You can open any drink page from the index and use the comparison and calculator tools to run exact arithmetic for the dataset numbers. For instance, open Cold brew, 200 mg, Espresso (double), 126 mg, Starbucks Pike Place Brewed (Grande), 310 mg, Dunkin' Original Blend (Medium), 210 mg, Red Bull (250 ml), 80 mg, Matcha (1 tsp), 70 mg, Bang, 300 mg and Club-Mate, 66 mg for concrete dataset examples and then press the tools button to use the calculators at the calculators. Each linked drink page is drawn from the open dataset and the calculators accept those exact mg values so the math you run is reproducible from the dataset.

If you want to audit provenance or download the whole table of 87 drinks, visit the dataset page at the open dataset and consult the licence at CC BY 4.0; the dataset page documents sources used for each entry and the licence allows reuse with attribution.

How do I report an error or suggest a new entry?

Report corrections or suggest new drink entries through our contact channel, which is documented on the site and links to the dataset update workflow. Use the corrections and contact page at corrections and contact and include the drink page URL and a primary source; we verify submissions against manufacturer or chain nutrition statements and update the dataset with attribution on the drink page.

Caveat: we accept community-sourced corrections but require a verifiable published source (manufacturer nutrition label, chain nutrition page, or lab report) before changing the dataset; this is the process described on the method page at method, sources and author and the dataset page at the open dataset documents changes and versioning under the CC BY 4.0 licence.

Where can I learn more about caffeine effects and limits?

For official intake benchmarks consult the U.S. Food and Drug Administration guidance, which lists 400 mg per day as a reference value for healthy adults at FDA, 400 mg/day for healthy adults. For sleep-timing evidence consult Drake et al. 2013 which measured sleep effects of 400 mg of caffeine given hours before bedtime at Drake et al. 2013.

For practical guides and comparisons on specific brands and products use our guides index at the guide index and specific articles such as the Bang energy guide at Bang Energy Drink: Exactly How Much Caffeine is Inside? and the espresso guide at Espresso Caffeine: How Much is in Your Shot? which link back to dataset numbers on the drink pages cited in those guides.

CoffeeLog mention: if you track cups with an external log such as CoffeeLog be careful to use the exact mg values from our dataset pages (for example Latte (large, 2 shots), 126 mg) when you add entries so your running totals match the dataset arithmetic.

End of explanatory hub text; every drink number in this page is taken from the open dataset and each statistic above cites the page that carries the numeric value in the same sentence. For tool use start at caffeine half-life calculator, latest cup before bedtime, safe daily limit by body weight and caffeine comparison.