Caffeine estimate methodology
The caffeine estimator gives a range, not a single number, because the caffeine in a cup varies with the coffee, the roast, the grind and the recipe, even between two cups made the same way. Nothing on CoffeeHQ measures the caffeine in your cup. Every figure the tool shows is a calculation, and this page sets out what it is calculated from.
The calculation
The caffeine in the drink is taken as the dose in grams multiplied by a range of milligrams per gram for the brewing method. That total is spread across the whole brew and your serving takes its share, so a small cup from a large batch gets a proportionally smaller estimate. Water added afterwards changes nothing: it dilutes the drink without removing caffeine.
| Method | Range the estimator uses | Span of the measurements below |
|---|---|---|
| Espresso | 6.0 to 11.0 mg per gram | 6.0 to 10.9 mg per gram |
| Hot filter or immersion | 5.5 to 10.5 mg per gram | 6.0 to 10.5 mg per gram |
| Cold brew | 7.0 to 10.5 mg per gram | 7.4 to 10.5 mg per gram |
| Instant coffee (per gram of powder) | 19.0 to 58.0 mg per gram | 19.2 to 57.6 mg per gram |
Each range is the span of the measurements in which the study says how much coffee was used, expressed as caffeine in the cup and rounded outward to the nearest half milligram. That rule is checked automatically, so a range here cannot drift away from the studies listed below.
The ranges for the three brewed methods are for arabica, because that is what the studies brewed. One more check is applied to them. A drink cannot hold more caffeine than the coffee it was made from, and a typical arabica bean is put at about 13 mg per gram in the earlier literature that one of the studies below quotes. Every arabica range here ends beneath that. It is a check that the measured brews are plausible and not a limit on your coffee: the same study measured arabicas well above it. Robusta is handled differently and has a section of its own below.
What changed, and why
Until October 2026 the estimator used 3.3 to 4.4 mg per gram for espresso, 5.3 to 6.7 for filter and immersion brews and 6.0 to 8.5 for cold brew. Those figures had been chosen to agree with the caffeine ranges on CoffeeHQ's drink pages, not taken from measurements. Set beside studies that state the dose, they were low: by about half for espresso, by about a quarter for filter coffee and by less for cold brew. An 18 g espresso that the old ranges put at 59 to 79 mg is 108 to 198 mg on the new ones, and the one study that weighed such shots measured the equivalent of 155 to 176 mg at a 1:2 ratio.
The ranges are wider than before because the measurements are. That is a more honest picture of how well a recipe predicts the caffeine in a cup.
The new ranges were then checked by a second reader against the papers' own tables. Every transcribed value matched. One thing was changed as a result: the filter and cold-brew ranges had been taken from the caffeine dissolved in all the water added, which is more than reaches the cup, and at the top it came to about as much caffeine as an arabica bean holds. They are now estimates of caffeine in the cup, as described below, which lowered the filter range from 6.5 to 11.5 mg per gram to 5.5 to 10.5, and the cold-brew range from 9.0 to 12.5 to 7.0 to 10.5. The espresso and instant ranges, which were measured in the cup, did not move.
The measurements
Two kinds of figure are kept apart. Where a study weighed the drink, caffeine per gram is the caffeine in the cup divided by the dose. Where a study reports only the concentration of the liquid and the water added, the figure is concentration multiplied by the water and divided by the dose: that is the caffeine dissolved, and it overstates what reaches the cup by whatever liquid the wet grounds keep. None of these studies reports how much liquid was recovered, so the step to the cup is an estimate: spent grounds are taken to keep one and a half to two and a half times their own weight in liquid, the same rule of thumb the cold-brew calculator uses, with the larger loss set against a study's lowest concentration and the smaller against its highest. The tables show both figures and say which is which. All of the arithmetic is CoffeeHQ's; the measured values are the studies'.
Espresso
| Study and preparation | Coffee | What was measured | Per gram of coffee used | How it was read |
|---|---|---|---|---|
| Schmieder et al. 2023: 20 g dose, first 20 g of espresso (1:1) | 100% arabica, washed, Colombia | 120.4 to 139.8 mg per serving | 6.0 to 7.0 mg per gram | measured in the drink; full text and tables. Table 2, caffeine column, brew ratio 1/1. |
| Schmieder et al. 2023: 20 g dose, first 40 g of espresso (1:2) | 100% arabica, washed, Colombia | 172.4 to 195.2 mg per serving | 8.6 to 9.8 mg per gram | measured in the drink; full text and tables. Table 2, caffeine column, brew ratio 1/2. The paper's own average for a 40 g cup is 4.57 mg per gram of drink, which is 183 mg. |
| Schmieder et al. 2023: 20 g dose, first 60 g of espresso (1:3) | 100% arabica, washed, Colombia | 189.6 to 217.2 mg per serving | 9.5 to 10.9 mg per gram | measured in the drink; full text and tables. Table 2, caffeine column, brew ratio 1/3. |
One study, one coffee. It shows clearly that a longer shot carries more caffeine from the same dose: about 6 to 7 mg per gram at 1:1, 8.6 to 9.8 at 1:2 and 9.5 to 10.9 at 1:3. The estimator uses the whole span for every espresso and does not narrow it by ratio, because one coffee is too little to build a rule on. A short shot will tend to sit at the low end of the estimate and a long one at the high end.
Hot filter and immersion
| Study and preparation | Coffee | What was measured | Per gram of coffee used | How it was read |
|---|---|---|---|---|
| Lindsey et al. 2024: AeroPress with paper filter, 10 g in 150 g water, 1, 2 and 10 minute brews, five roast degrees | arabica, two Ethiopian coffees (one natural, one washed) | 566.0 to 697.0 mg per litre | 7.1 to 9.4 mg per gram in the cup (estimate); 8.5 to 10.5 dissolved | concentration x water added; full text and tables. Results, "Caffeine content": 134-165 and 145-165 mg per 8 oz cup. Reported per 8 fl oz cup; converted here at 236.6 ml to the cup. |
| Fuller and Rao 2017: French press, 35 g in 350 ml water at 98 °C, 6 minutes | arabica (Kona Typica), medium and dark roast, medium and coarse grind | 840.0 to 1060.0 mg per litre | 6.3 to 9.0 mg per gram in the cup (estimate); 8.4 to 10.6 dissolved | concentration x water added; full text and tables. Table 3, hot brew caffeine concentration. |
| Jeon et al. 2019: paper filter, about 10 g with 200 ml boiling water in three pours, 2 minutes | arabica, 11 single-origin supermarket coffees | 342.0 to 566.0 mg per litre | 6.0 to 10.5 mg per gram in the cup (estimate); 6.8 to 11.3 dissolved | concentration x water added; full text and tables. Table 4, rows RG 1-11, caffeine per serving of 200 ml (mean 95.0). |
| Heo et al. 2020: drip coffee maker with paper filter, 7 g per 100 ml water | a blend, a Colombian and a Kenyan coffee, species not stated | 740.0 to 770.0 mg per litre | 8.7 to 9.8 mg per gram in the cup (estimate); 10.6 to 11.0 dissolved | concentration x water added; full text and tables. Table 3, BlendingCM, ColombiaCM, KenyaCM. A fourth coffee, a supermarket ground coffee of unstated species, measured 1,050 mg per litre and is left out of this row. |
| Wierzejska and Gielecińska 2024: one teaspoon (2.5 g) of ground coffee steeped 5 minutes in 160 ml hot water, then decanted | arabica (mean of the arabica samples) | 18.1 mg per serving | 7.2 mg per gram | measured in the drink; full text and tables. Results, section 3.1: arabica mean 18.1 mg per serving against 25.8 mg for blends. A mean, not a range. Across all 65 ground coffees, blends included, a serving held 12.8 to 45.6 mg. |
Cold brew
| Study and preparation | Coffee | What was measured | Per gram of coffee used | How it was read |
|---|---|---|---|---|
| Fuller and Rao 2017: 35 g in 350 ml water at 21-25 °C, sampled to 24 hours | arabica (Kona Typica), medium and dark roast, medium and coarse grind | 990.0 to 1230.0 mg per litre | 7.4 to 10.5 mg per gram in the cup (estimate); 9.9 to 12.3 dissolved | concentration x water added; full text and tables. Table 3, cold brew caffeine concentration after 1,440 minutes. Same coffees and ratio as the French press row above. |
| Heo et al. 2020: 7 g per 100 ml water, steeped 9 hours at 4 °C | a blend, a Colombian and a Kenyan coffee, species not stated | 630.0 to 800.0 mg per litre | 7.4 to 10.2 mg per gram in the cup (estimate); 9.0 to 11.4 dissolved | concentration x water added; full text and tables. Table 3, BlendingCB, ColombiaCB, KenyaCB. The supermarket ground coffee measured 950 mg per litre and is left out of this row. |
The cold-brew range starts higher than the hot one and ends at the same place. In the study that brewed one coffee both ways at the same ratio, the 24-hour cold steep held more caffeine than the six-minute French press in all four coffees, clearly so in three. In the other, a nine-hour steep in a refrigerator did not differ consistently from a drip machine. A long steep can dissolve as much caffeine as hot water and sometimes more; where the cold-brew estimate is higher it reflects the length of the steep, not the cold. Two studies are too few to say more.
Instant coffee
| Study and preparation | Coffee | What was measured | Per gram of coffee used | How it was read |
|---|---|---|---|---|
| Wierzejska and Gielecińska 2024: one teaspoon (2.0 g) of instant coffee dissolved in 160 ml hot water | not stated | 38.4 to 115.2 mg per serving | 19.2 to 57.6 mg per gram | measured in the drink; full text and tables. Table 3, instant coffee: 9.6-28.8 % of 400 mg per serving (mean 16.0 %). The paper prints the range as a share of 400 mg; milligrams here are that share multiplied out. The mean is 64 mg. |
| Jeon et al. 2019: one stick of 100% instant coffee, labelled as about 1 g, dissolved in 100 ml | not stated | 23.7 to 71.2 mg per serving | not derivable | per serving; dose not stated; full text and tables. Table 2, rows I 1-11, caffeine per serving of 100 ml: 23.7 to 71.2 mg, mean 37.0 ± 14.5. Per stick. The paper gives the stick as approximately 1 g from the label and did not weigh it, so no per-gram figure is derived: 71 mg from exactly one gram would be 7 per cent caffeine. |
| USDA FoodData Central, FDC 171893: instant coffee powder, per gram | not stated | 31.4 mg per serving | 31.4 mg per gram | measured in the drink; dataset record. Beverages, coffee, instant, regular, powder: caffeine 3,140 mg per 100 g. |
For instant coffee the dose is the powder itself, so the figure is caffeine per gram of powder, not per gram of the coffee it was made from. Products differ threefold, which is why this is the widest range the tool gives. The range rests on one study of 26 products made with a 2.0 g teaspoon; a food-composition record of 31 mg per gram sits inside it. A second study measured 24 to 71 mg from single-serve sticks, but gives the stick only as about one gram from the label, so no figure per gram is taken from it.
Coffees bought ready-made
| Study and preparation | Coffee | What was measured | Per gram of coffee used | How it was read |
|---|---|---|---|---|
| USDA FoodData Central, FDC 171890: brewed coffee, prepared with tap water | not stated | 400.0 mg per litre | not derivable | per serving; dose not stated; dataset record. Caffeine 40 mg per 100 g, derivation: analytical. The record does not say how much coffee was used. |
| USDA FoodData Central, FDC 171891: espresso, restaurant-prepared | not stated | 2120.0 mg per litre | not derivable | per serving; dose not stated; dataset record. Caffeine 212 mg per 100 g, derivation: analytical. The record does not say how much coffee was used. |
| Wierzejska and Gielecińska 2024: espresso bought from shops, mean serving 34.1 ml (16-60 ml) | not stated | 32.5 to 153.0 mg per serving | not derivable | per serving; dose not stated; full text and tables. Table 1, espresso column, minimum and maximum across all outlets; Table 3 mean 18.8 % of 400 mg. The mean is 75 mg. The authors say they had no information on how the shops prepared the coffee. |
| Crozier et al. 2012: 20 espressos bought from coffee shops | not stated | 322.0 mg per serving | not derivable | per serving; dose not stated; abstract only. Abstract: six-fold differences in caffeine; the most in a single espresso was 322 mg. The highest value only. The abstract gives a six-fold range and no lowest figure, and the paper itself was not opened. |
| Ludwig et al. 2014: 104 espressos bought from coffee shops in Scotland, Italy and Spain | not stated | 48.0 to 317.0 mg per serving | not derivable | per serving; dose not stated; abstract only. Abstract: caffeine ranged from 48 to 317 mg per serving. |
| McCusker, Goldberger and Cone 2003: brewed specialty coffees bought from coffee shops, 16 US fl oz cups | not stated | 188.0 mg per serving | not derivable | per serving; dose not stated; abstract only. Abstract: mean (SD) 188 (36) mg for a 16-oz cup. A mean. The abstract also reports 259 to 564 mg for one drink bought from one outlet on six days. |
None of these says how much coffee went into the drink, so none can be turned into a figure per gram and none sets a range in the tool. They are here because they show how far real servings spread, and as a check on the tool's output.
The check is rough, because the doses are unknown. The survey of 58 shop espressos found a mean of 75 mg in a serving of about 34 ml. If those were single shots made from 7 to 9 g of coffee, the estimator would put them at 42 to 99 mg, which contains the survey's mean. For a double made from 18 g it gives 108 to 198 mg; the two larger surveys, read only as abstracts, found shop espressos from 48 to 317 mg and one of 322 mg. For brewed coffee, a 16 fl oz cup measured at a mean of 188 mg would be estimated at 143 to 273 mg if it was made with 55 g of coffee to the litre, which is an assumption and not something the study reports. Nothing here contradicts the ranges; nothing here proves them either.
Robusta
Every dose-stated measurement above was made with arabica, or with coffee whose species is not given. CoffeeHQ has read no study that brewed robusta, stated the dose and measured the caffeine in the cup. So the estimator cannot give robusta a measured range. What it gives instead is a bound, worked from what the bean holds.
| Method | Robusta: the bound the estimator uses | Arabica: the measured range |
|---|---|---|
| Espresso | 6.0 to 41.0 mg per gram | 6.0 to 11.0 mg per gram |
| Hot filter or immersion | 5.5 to 41.0 mg per gram | 5.5 to 10.5 mg per gram |
| Cold brew | 7.0 to 41.0 mg per gram | 7.0 to 10.5 mg per gram |
The top, 41.0 mg per gram, is the caffeine in the richest robusta measured in the papers below, as if all of it reached the cup. That coffee was one green robusta of 42 from Java, at 3.53 per cent of its dry weight, which is 35.3 mg per gram. Roasted coffee weighs less than the green coffee it came from, so the same caffeine is a larger share of each gram: with green coffee at 8 per cent moisture and a roast that loses 20 per cent of its weight, and no caffeine lost in the roaster, 35.3 becomes 40.6, which is rounded up. The moisture and the roast loss are not that study's. They are the ends of the ranges that CoffeeHQ's green coffee and roasting pages report, taken in the direction of more caffeine. The highest figure measured on roasted robusta directly, 33.29 mg per gram, lies beneath it.
No brew takes everything out of the grounds, and wet grounds keep some of the liquid, so a real cup holds less than this. How much less has not been measured for robusta. The top of the range is therefore a margin of caution, not what a cup of robusta is expected to contain.
The bottom is the bottom of the arabica range for the same method. The two species overlap in both studies that measured the two. Among the 74 green coffees the lowest robusta, at 1.95 per cent, held less than the highest arabica, at 2.37. The one roasted robusta measured ran from 33.29 down to 9.27 mg per gram across six roasts, while the espresso study above drew as much as 10.9 mg per gram into the cup from an arabica. Nothing read rules out a robusta cup as low as an arabica one. Most will be higher: the 42 robustas averaged 2.9 per cent against 1.8 for the 32 arabicas.
So an 18 g espresso reads 108 to 198 mg as arabica. For robusta the estimator does not print one range. It makes two separate statements: at least as an arabica brewed the same way, 108 to 198 mg (measured brews); and cannot be more than about 738 mg, which is everything in the richest robusta measured, if all of it reached the cup. No study read measured brewed robusta with a stated dose, so where between the two a given robusta falls is not known. The two figures are not the ends of one estimate, and the ceiling does not mean robusta carries that many times the caffeine of arabica.
| Study and sample | Coffee | Caffeine in the bean | Per gram once roasted (estimate) | How it was read |
|---|---|---|---|---|
| Adnan et al. 2020: mean of 32 green arabicas from Java, 2014 harvest | Arabica, green | 18.0 mg per gram of dry weight | 18.3 to 20.7 mg per gram | UV-Vis after extraction into dichloromethane, in duplicate. Results 3.1.1: "Arabica had a caffeine content of 1.8% dw". |
| Adnan et al. 2020: lowest and highest of the 32 arabicas as Figure 1a plots them | Arabica, green | 12.0 to 23.7 mg per gram of dry weight | 12.2 to 27.3 mg per gram | UV-Vis after extraction into dichloromethane, in duplicate. Figure 1a and supplementary Table S3 (1.20 and 2.37 per cent). Table S3 labels four samples of 2.72 to 3.21 per cent as arabica and four of 1.67 to 1.88 as robusta. Figure 1a plots no arabica above 2.37 and no robusta below 1.95, and the table gives the means in the text only if at least those eight labels are exchanged, so the figure is followed here. A second reading found that exchanging those eight and two more (a 2.36 labelled arabica and a 2.00 labelled robusta) reproduces the test the paper prints for the difference between the species, which no other set tried does. |
| Adnan et al. 2020: mean of 42 green robustas from Java, 2014 harvest | Robusta, green | 29.0 mg per gram of dry weight | 29.5 to 33.3 mg per gram | UV-Vis after extraction into dichloromethane, in duplicate. Results 3.1.1: "while Robusta contained 2.9% dw". |
| Adnan et al. 2020: lowest and highest of the 42 robustas as Figure 1a plots them | Robusta, green | 19.5 to 35.3 mg per gram of dry weight | 19.8 to 40.6 mg per gram | UV-Vis after extraction into dichloromethane, in duplicate. Figure 1a and supplementary Table S3 (1.95 per cent at location 24; 3.53 per cent at location 11, variety BP534). 3.53 per cent is the highest single value in the study under either labelling. See the arabica row for the labels in Table S3. |
| Freitas et al. 2023: one dry-processed Coffea canephora lot from Minas Gerais, six roast profiles, each roasted once and analysed in triplicate | Robusta, roasted | 9.3 to 33.3 mg per gram | as measured | HPLC with diode-array detection. Table 3, caffeine, C. canephora row: 33.29, 31.16, 19.27, 12.40, 9.27 and 30.19 mg/g. Per gram of roasted ground coffee; the paper does not say whether on a dry basis. The highest figure is the gentlest roast (135 °C for 20.20 minutes) and the lowest the harshest (275 °C for 7.46 minutes). The paper calls caffeine heat-stable and does not explain a fall of this size. |
| Freitas et al. 2023: one dry-processed Catuaí arabica lot from Minas Gerais, six roast profiles, each roasted once and analysed in triplicate | Arabica, roasted | 5.6 to 10.2 mg per gram | as measured | HPLC with diode-array detection. Table 3, caffeine, C. arabica row: 5.55, 6.71, 10.15, 6.19, 9.33 and 7.19 mg/g. Per gram of roasted ground coffee; the paper does not say whether on a dry basis. By the paper's own test the six arabica figures do not differ from one another. |
| Weldegebreal et al. 2017: one green coffee bought at a local market, variety and origin not recorded, measured by four methods in triplicate | Species not stated, green | 14.0 to 15.2 mg per gram | not worked | FT-IR-ATR, NIR, fluorescence and UV-Vis. Table 1, mean column: 1.520, 1.500, 1.497 and 1.397 per cent w/w. A methods paper. It compares its result with published arabica figures but does not state the species of its sample, or whether the percentage is of dry weight. |
| Where it is quoted, and from what | Coffee | Caffeine in the bean | Per gram once roasted (estimate) | How it was read |
|---|---|---|---|---|
| quoted by Weldegebreal et al. 2017: a range from a 2015 review of analytical methods, the paper's reference 5 | Arabica, green | 7.0 to 16.0 mg per gram | not worked | not given. Background: "Green coffee beans of Coffea arabica contains between 0.7 and 1.6% caffeine and of Coffea robusta between 1.5 and 4.0% caffeine [5]". |
| quoted by Weldegebreal et al. 2017: a range from a 2015 review of analytical methods, the paper's reference 5 | Robusta, green | 15.0 to 40.0 mg per gram | not worked | not given. Background, the same sentence. The top of this range, 4.0 per cent, is above every figure measured in the papers read. It is second-hand and sets nothing in the tool. |
| quoted by Adnan et al. 2020: one figure from a reference work, the paper's reference 16; arabica 1.3 per cent | Robusta, green | 23.0 mg per gram of dry weight | not worked | not given. Introduction: "the caffeine (1.3% dw) ... content in Arabica are lower than in Robusta (2.3% dw ...) [16]". |
| quoted by Adnan et al. 2020: the range in another study, the paper's reference 31 (mean 1.8 per cent); arabica 0.8 to 1.8 per cent, mean 1.3 | Robusta, green | 12.0 to 25.0 mg per gram of dry weight | not worked | not given. Discussion 4.1.1: "ranged from 0.8% to 1.8% dw (mean = 1.3% dw) in Arabica as compared with a range of 1.2% to 2.5% dw in Robusta (mean = 1.8% dw) [31]". |
These are few coffees and they do not describe either species as a whole: 74 green coffees from one island in one harvest, one roasted lot of each species from one town, and one green coffee from a market. The quoted figures were not opened and set nothing in the tool. One of them puts green robusta as high as 4 per cent, above anything measured here, so the top of the range is not a guarantee either.
Until October 2026 the estimator multiplied the arabica range by 1.5 to 2.0 for robusta. That was replaced for two reasons. The papers do not support one ratio: the averages of the 74 green coffees stand at about 1.6 to one, while the two roasted lots stand anywhere from level to six to one depending on the roast. And the most the old rule allowed a gram of robusta to give was 22 mg, which is below the average of the robustas measured. The old rule put an 18 g robusta espresso at 162 to 396 mg.
The robusta range is not applied to instant coffee, whose figures are for the powder as sold. If you leave the species unset the arabica range is used; a blend with robusta in it can hold a good deal more.
Decaf
| Study and preparation | Coffee | What was measured | Per gram of coffee used | How it was read |
|---|---|---|---|---|
| McCusker, Goldberger and Cone 2003: coffees sold as decaffeinated, bought from coffee shops | not stated | 17.7 mg per serving | not derivable | per serving; dose not stated; abstract only. Abstract: decaffeinated coffees had less than 17.7 mg per dose. An upper figure only; the abstract gives no lowest value. |
| McCusker et al. 2006: decaffeinated brewed coffee from ten outlets, 16 US fl oz servings | not stated | 0.0 to 13.9 mg per serving | not derivable | per serving; dose not stated; abstract only. Abstract: 0-13.9 mg per 16-oz serving. |
| McCusker et al. 2006: decaffeinated espresso shots from one outlet | not stated | 3.0 to 15.8 mg per serving | not derivable | per serving; dose not stated; abstract only. Abstract: 3.0-15.8 mg per shot. |
Decaf is never treated as free of caffeine. The estimator shows 0 to 18 mg for a serving whatever you enter, which is the span of the shop-bought decaf above, rounded up. All three figures were read from abstracts only, and none gives a basis for scaling the figure with the dose, so the tool marks it as inferred. A further study, read in full, found caffeine below its limit of quantification in two decaffeinated instant coffees and one decaffeinated ground coffee.
What the estimate cannot do
- It cannot see your coffee. Two arabicas can differ in caffeine before either is brewed.
- It rests on few studies: one for espresso, four for hot brews, two for cold brew. More measurements could move a range.
- It has no measured range for brewed robusta. What it shows for robusta is the arabica range as a lower statement and, separately, a ceiling from the caffeine in the bean.
- The arabica ranges are the span of the coffees that were brewed. An arabica richer in caffeine than those can brew above them.
- For filter and cold brew the step from caffeine dissolved to caffeine in the cup is an estimate of how much liquid the grounds keep, not something those studies measured.
- It does not narrow the espresso range for a short or a long shot.
- It has no figures for moka pots, capsules, Turkish coffee or ready-to-drink cans.
- It estimates the caffeine from the coffee and nothing else. It is not a day's total: tea, chocolate, energy drinks and other products add their own. Where a product states its caffeine per serving, that figure is counted separately. A plant ingredient on a label, such as guarana, is weighed as the ingredient, not as caffeine, and where a product does not state its caffeine the amount is unknown rather than zero. The caffeine guide sets out why.
What this is not
This estimator is not medical advice. It does not know your personal caffeine tolerance, does not account for medication interactions, and makes no claim about what's safe during pregnancy or for any health condition. If caffeine intake matters for your health, speak to a healthcare professional rather than relying on this or any online estimate.
Sources
- B. K. L. Schmieder, V. B. Pannusch, L. Vannieuwenhuyse, H. Briesen, M. Minceva. Influence of Flow Rate, Particle Size, and Temperature on Espresso Extraction Kinetics. Foods (MDPI).
- Z. R. Lindsey, J. R. Williams, J. S. Burgess, N. T. Moore and P. M. Splichal. Caffeine content in filter coffee brews as a function of degree of roast and extraction yield. Scientific Reports, 14, 29126.
- M. Fuller and N. Z. Rao. The Effect of Time, Roasting Temperature, and Grind Size on Caffeine and Chlorogenic Acid Concentrations in Cold Brew Coffee. Scientific Reports, 7, 17979.
- J.-S. Jeon, H.-T. Kim, I.-H. Jeong, S.-R. Hong, M.-S. Oh, M.-H. Yoon, J.-H. Shim, J. H. Jeong and A. M. Abd El-Aty. Contents of chlorogenic acids and caffeine in various coffee-related products. Journal of Advanced Research, 17, 85-94.
- J. Heo, K. Adhikari, K. S. Choi and J. Lee. Analysis of Caffeine, Chlorogenic Acid, Trigonelline, and Volatile Compounds in Cold Brew Coffee Using High-Performance Liquid Chromatography and Solid-Phase Microextraction-Gas Chromatography-Mass Spectrometry. Foods, 9(12), 1746.
- R. E. Wierzejska, I. Gielecińska. Evaluation of the Caffeine Content in Servings of Popular Coffees in Terms of Its Safe Intake. Nutrients, 16(15), 2385.
- U.S. Department of Agriculture (USDA). FoodData Central caffeine content entries.
- T. W. M. Crozier, A. Stalmach, M. E. J. Lean and A. Crozier. Espresso coffees, caffeine and chlorogenic acid intake: potential health implications. Food & Function, 3, 30-33.
- I. A. Ludwig, P. Mena, L. Calani, C. Cid, D. Del Rio, M. E. J. Lean and A. Crozier. Variations in caffeine and chlorogenic acid contents of coffees: what are we drinking?. Food & Function, 5, 1718-1726.
- R. R. McCusker, B. A. Goldberger and E. J. Cone. Caffeine content of specialty coffees. Journal of Analytical Toxicology, 27(7), 520-522.
- R. R. McCusker, B. Fuehrlein, B. A. Goldberger, M. S. Gold and E. J. Cone. Caffeine content of decaffeinated coffee. Journal of Analytical Toxicology, 30(8), 611-613.
- A. Adnan, M. Naumann, D. Mörlein and E. Pawelzik. Reliable discrimination of green coffee beans species: a comparison of UV-Vis-based determination of caffeine and chlorogenic acid with non-targeted near-infrared spectroscopy. Foods, 9(6), 788.
- V. V. Freitas, L. L. Rodrigues Borges, G. A. Dias Castro, M. H. dos Santos, M. C. T. R. Vidigal, S. A. Fernandes and P. C. Stringheta. Impact of different roasting conditions on the chemical composition, antioxidant activities, and color of Coffea canephora and Coffea arabica L. samples. Heliyon, 9(9), e19580.
- B. Weldegebreal, M. Redi-Abshiro and B. S. Chandravanshi. Development of new analytical methods for the determination of caffeine content in aqueous solution of green coffee beans. Chemistry Central Journal, 11, 126.
Last reviewed October 2026.