How much caffeine is in tea?
A cup of tea generally lands somewhere between about 15 and 60 mg of caffeine. A cup of coffee generally lands between about 75 and 165 mg. Both of those ranges are wide, and they are wide for the same reason: how you make it matters more than what you bought.
I have been going to China for thirty-odd years, I drink something like twelve to sixteen cups a day (I stop at 3:30, which is less a rule than a hard-won piece of self-knowledge), and in all that time the question I get asked more than any other is which of our teas has the most caffeine in it. For years this page answered with a tidy little table of milligrams. This autumn we went and checked that table against the published science, and I am sorry to report that it did not survive the experience.
So here is the honest version, which I think is more useful anyway.
Why the tidy table had to go
The comfortable story is that the six types of tea line up in a neat row: white at the bottom, then green, then oolong, then black at the top. Nearly every tea website prints some version of it. The trouble is that the six types are processing categories. White, green, oolong, black and pu-er differ by how the leaf was withered, oxidised, fixed and fired. And it turns out processing is very nearly the last thing that touches caffeine.
A team at National Taiwan University, working with Taiwan's Tea and Beverage Research Station, followed nine teas through every stage of manufacture and measured the caffeine at each one. Across all of them the caffeine moved less than 20%, with no relationship whatsoever to how oxidised the tea became. Their conclusion, in their words: "plucking maturity, rather than manufacturing steps, is the dominant determinant of caffeine levels."
Which is to say: what you did to the leaf barely matters. Which leaf you picked matters enormously. Tony Gebely, whose account of tea chemistry I have leaned on for years, lists the things that actually drive caffeine as the cultivar, the climate, the age of the leaf and how the plant was propagated. Tea type is not on his list. It never was. I simply had not noticed.
What actually changes the caffeine in your cup
Here is the same question answered properly, biggest lever first.
| How long you steep it | Thirty seconds pulls about a third of the leaf's caffeine out. Five minutes pulls better than nine tenths of it. |
| How hot the water is | One oolong measured 4 mg at 80 °C and 14 mg at 100 °C, same leaf, same two minutes. |
| Loose leaf or a bag | Brewed identically, black tea in a cup-size bag came out at 39 mg against 24 mg loose. |
| Which plant it came from | Caffeine can swing roughly tenfold between cultivars of the same species. |
| Which leaves got picked | The bud runs about twice the caffeine of the fifth leaf down the stem. |
| Where it grew | Caffeine tends to run lower at higher elevation, and shade pushes it up. |
| Which of the six types it is | Last. Processing moves caffeine less than 20%, and the average gap between whole categories is smaller than the spread inside any one of them. |
If you want less caffeine in your cup this afternoon, the lever is the first row, not the last one. Use a little less leaf, keep the water off the boil, steep it briefly and pour it off. That will do more than switching teas ever will.
What the measurements actually look like
The best set of numbers I know of comes from Tea: History, Terroirs, Varieties by Kevin Gascoyne, François Marchand, Jasmin Desharnais and Hugo Américi (Firefly Books, second edition, 2014). They commissioned a laboratory to run 31 teas by liquid chromatography, each one brewed the way that tea is actually supposed to be brewed, and reported the caffeine per cup. That last part is rarer than it sounds and it is why their figures are worth more than most.
Grouping their teas into families gives this. The numbers are theirs; the grouping is ours.
| Camellia Flowers | very little (see below) |
| White | 15–39 mg per cup |
| Oolong | 12–49 mg per cup |
| Green | 12–50 mg per cup |
| Pu-er | 19–23 mg per cup |
| Black | 16–58 mg per cup |
The point of that table is that it has no order. Every range overlaps every other range. There is no line you can draw across it that puts one type above another. Their own summary of the exercise: "there is no real consistency in this area. In every tea family, some teas contain very little caffeine while others contain a lot." Do read it as six overlapping puddles rather than a ladder, and bear in mind these were their teas, not ours, with only a handful in each family.
Where I was wrong about oolong
When we first went through this research I said, flatly, that oolong belongs near the top of the list. I have been drinking the stuff for decades and I know what our oolongs do to me.
It turns out I was half right, which is the most annoying possible outcome.
In every large survey, oolong has the lowest category average of any tea type. A Chinese study of 1,398 teas from seventeen provinces put it bottom of seven. A study of thirty named Chinese teas put Tieguanyin second-lowest of the whole thirty. So as a category I was wrong.
But Gascoyne's laboratory measured Bai Hao oolong at 49 mg per cup, fifth highest of all 31 teas they tested, and Bai Hao is the style we sell as Eastern Beauty. Meanwhile the Wuyi rock oolongs, the family our Da Hong Pao belongs to, measured down at 16 mg. That is a threefold spread inside one category, on one shelf, in one shop.
So I was right about the oolong I reach for and wrong about oolong as a class, and both of those things can be true at once because the class was never the thing doing the work. I find this genuinely more interesting than the table I had before. (I should say plainly that those were their teas and not ours. Nobody has put a Little Red Cup oolong through a chromatograph, so treat 49 mg as "this style measured high," not as a reading off our own tin.)
The one number we did measure
There is exactly one caffeine figure on this page that came out of a laboratory we paid, so I may as well show you the paperwork. In September 2014 we sent one ounce of our Camellia Flowers to Green Mountain Laboratories in Montpelier, Vermont, with a handwritten note asking them to test for caffeine and tell us the amount, if any. They ran it by HPLC with UV detection. The answer came back 0.516% by weight, on certificate #622E, and that is the only caffeine number on this page that anybody has ever measured on our own tea.
Tea leaves generally run somewhere between two and five percent. So the flowers carry roughly a fifth to a quarter of what the leaf does, and a cup of them is a very gentle thing indeed.
I want to be careful about what that number is, because being careful is the whole point of this rewrite. It is a measurement of the dried flowers by weight, from one sample, eleven years ago. It is not the same claim as a milligram figure for your cup, and I am not going to pretend otherwise. What I can tell you is that it is low, that it is ours rather than one we borrowed, and that an unrelated laboratory testing Azorean tea flowers landed at 0.615%. Two labs, two continents, a decade apart, and the same answer to within a rounding step. I find that fairly convincing.
We used to say the flowers were about the same as decaffeinated tea. That was wrong, and our own certificate is what shows it: properly decaffeinated tea works out somewhere around a tenth of a percent by weight, so at 0.516% the flowers sit several times above it, not below. The flowers are low. They are not decaf. I would rather tell you that than keep a tidier sentence.
Does white tea have caffeine?
Yes, and rather more than its gentle reputation suggests. White tea is barely processed, and since processing is not what puts caffeine in the leaf, being barely processed does not make it low. It is often made from buds, and buds are the most caffeinated part of the plant.
How much caffeine is in white tea?
Measured per cup, white teas have come in between about 15 and 39 mg. A Bai Mu Dan, the style of our White Peony, measured 39 mg in Gascoyne's analysis, which put it above every oolong and every pu-er on the same chart. By dry weight, white tea ranks fourth of six categories. It is a middling tea for caffeine, not a low one.
Which type of tea has the least caffeine?
No type reliably has the least, and anyone who tells you otherwise is repeating a chart rather than a measurement. White is usually nominated and the measurements do not support it. If you want the least caffeine from a real tea, brew it briefly and off the boil. If you want next to none, drink the flowers.
Does black tea have more caffeine than other types of tea?
On average, slightly, and the average conceals almost everything worth knowing. Black teas have been measured from 16 to 58 mg per cup, which overlaps greens, oolongs and whites almost completely. A strong green will beat a mild black without difficulty.
Does tea have more caffeine than coffee?
No, and it is not close. The FDA puts a 12-ounce cup of ordinary brewed coffee at 113 to 247 mg, which is roughly 75 to 165 mg for an 8-ounce cup. Even a strong black tea sits well below the bottom of that.
Coffee's own number is no steadier than tea's, incidentally. When researchers bought the same coffee from the same shop on six consecutive days, it ran from 259 to 564 mg per 16 ounces. Same drink, same counter, more than double. (We used to say on this page that light roasts carry more caffeine than dark. Two controlled studies since found no meaningful difference, so that has come out too.)
Does the brewing process affect the caffeine content of tea?
Enormously, and it is the part you control. Thirty seconds of steeping extracts roughly a third of the leaf's caffeine; five minutes gets you past ninety percent. Water temperature and how much leaf you use move it further. Brew our White Peony lightly and pour it over ice, which makes an utterly splendid iced tea, and the caffeine in the glass is genuinely trivial.
Does rinsing the leaves wash the caffeine out?
No. This one is worth killing properly, because it is everywhere. The idea is that a quick first steep, thrown away, takes most of the caffeine with it. In practice a thirty-second rinse removes something like a tenth to a third of it, and Gascoyne's people found a second infusion of the same leaves still carried 86% as much caffeine as the first. You are pouring your best infusion down the sink for very little.
Is decaffeinated tea completely caffeine-free?
No. The FDA's own heading on the subject is "'Decaffeinated' Does Not Mean Caffeine-free," and measured decaffeinated teas come in under 12 mg per serving rather than at zero. We do not carry any regardless. We are a small, family-operated company and low-volume decaffeination done well is not economically feasible for us. Sorry.
Does tea contain other compounds that interact with caffeine?
Yes, and one of them is why a cup of tea feels different from a cup of coffee carrying the same caffeine.
- L-theanine, an amino acid found largely in tea, works against the jangling edge of caffeine while the alertness stays. Drinkers describe steadier concentration over a longer stretch than caffeine alone gives them.
- EGCG, the most abundant of tea's catechins, is the compound most of the antioxidant talk is about.
Tea holds a slew of others, and the honest position is that the interactions between them are not well enough pinned down for me to make you promises about any of it.
Are there any caffeine-free herbal teas?
Our Camellia Flowers are the closest thing we sell. They are not technically tea, being the blossoms rather than the leaves, though they do come from the tea plant itself. Very low caffeine, as above, and a lovely floral cup.
How does chamomile compare to other types of tea in terms of caffeine content?
Chamomile is a tisane and contains no caffeine at all, having nothing to do with the tea plant. We grow it in the garden. We do not sell it.
How much is too much
The FDA puts 400 mg a day, which it reckons at two to three 12-ounce cups of coffee, as an amount not generally associated with ill effects in healthy adults. Above that, your own sensitivity matters more than anybody's average. Mine evidently runs high, given the twelve to sixteen cups, though I would not hold myself up as a clinical benchmark.
Shop by caffeine level
With all the above ringing in your ears, here is roughly how I would steer you in person. These are tendencies, not readings.
For energy without the coffee crash
- Da Hong Pao Oolong, a rock-mineral oolong with cinnamon and caramel notes that brews strong without turning bitter.
- Our black teas, full-bodied, including Keemun, Yunnan Black and Russian Caravan.
To cut back without quitting
- White Peony, gentle and naturally sweet, and forgiving of a short steep.
- Our green teas, where Dragon Well and Jasmine are good places to start.
For as little as we can offer
- Camellia Flowers, the blossoms of the tea plant rather than its leaves. All the floral character, and the only caffeine number on this page we had measured ourselves.
Or browse the full catalog. Every tea page lists brewing notes, which, as you now know, is the part that actually decides what ends up in your cup.
Where these numbers come from
We rebuilt this page in September 2026 against the published literature. The sources doing the heavy lifting:
- Hung, Kuo et al., "Unveiling Fermentation Effects on the Functional Composition of Taiwanese Native Teas," Molecules 2026;31(1):171 — processing barely moves caffeine.
- Green Mountain Laboratories, Inc., Montpelier, Vermont — Certificate of Analysis #622E, sample received 12 September 2014, analysed 16 September 2014: our Camellia Flowers at 0.516% caffeine by weight, by HPLC with UV detection.
- Gascoyne, Marchand, Desharnais & Américi, Tea: History, Terroirs, Varieties, 2nd ed., Firefly Books, 2014 — the commissioned per-cup analysis of 31 teas that the family ranges above are derived from.
- Tang et al., "Phytochemical Composition and Antioxidant Capacity of 30 Chinese Teas," Antioxidants 2019;8(6):180 — thirty named Chinese teas by dry weight.
- Rana et al., J Food Sci Technol 2016;53(11):4023 — caffeine by leaf position.
- Ayakdaş & Ağagündüz, Foods 2025 — brewing temperature and time across six tea types.
- Paiva et al., Antioxidants 2021;10(2):183 — tea flowers against leaves.
- Astill et al., J Agric Food Chem 2001;49(11):5340 — brew-time extraction curves. Chin et al., J Anal Toxicol 2008;32(8):702 — the 14 to 61 mg per serving range. Yong et al., Food Chem Toxicol 2022;165:113102 — 1,398 Chinese teas.
- FDA, "Spilling the Beans: How Much Caffeine is Too Much" and Tony Gebely's tea chemistry.
If you spot something here that is wrong, I would genuinely like to know. That is how this page got better.
Mike
