Yes, You Can Brew Tea With Boiling Water

Tea Science｜Author：祥興茶行｜Published：2026-08-17

The claim that boiling water "destroys theanine" has circulated for years, yet no study has ever measured theanine going down. We went through the research, and for Taiwan's most-loved teas, the boiling-water cup tasted best.

Someone often says it while you're brewing: don't let the water reach a full boil, or the heat will knock out the theanine and the tea won't be sweet.

It sounds reasonable. Boiling water is very hot, tea leaves are tender, and everyone has tasted it for themselves: the same bag of green tea comes out more bitter the hotter the water. But after working through every study we could find, not one of them has measured theanine going down, for all the years this claim has been around.

The short answer: you can brew tea with boiling water. Taiwan's Tea Research and Extension Station brewed and tasted six Taiwanese teas one by one, and for high-mountain tea and Dong Ding oolong, the boiling-water cup tasted best [1]. The right temperature depends on which tea you have and which flavor you want from it.

The chart below puts two sets of official recommendations on one temperature axis: the further left, the more tender the bud; the further right, the more roasted, rolled, and oxidized the tea.

〔圖〕The brewing-temperature spectrum: teas arranged on one temperature axis by recommended water temperature, tender-plucked teas at the left end and roasted, rolled, and fully oxidized teas at the right; green markers are Taiwan’s Tea Research and Extension Station, gold markers are Japan’s Ministry of Agriculture, Forestry and Fisheries

Temperatures from the brewing manual of Japan's Ministry of Agriculture, Forestry and Fisheries (茶の淹れ方マニュアル) [18] and the Taiwan Tea Research Bulletin 31 (2012) [1]. The TRES half is a descriptive tasting conclusion: the study reports no panel size or scoring system, and no statistical test. For bancha, genmaicha, and hojicha the original says "熱湯" (rendered "boiling water" in the official English version); the manual gives no numeric temperature for these teas. Gyokuro's 50–60℃ only works with a very concentrated ratio: about 10 ml per cup, steeped 150 seconds. These temperatures are a starting point; adjust them to what you taste.

Parts 1–5 look at where the claim comes from and what actually happens in the cup; part 6 answers "what temperature should my tea get"; the last two parts cover "hard on the stomach," "destroys the vitamins," and what people a thousand years ago were worried about.

1. Two claims: heat broke it down, or it was drowned out

Theanine is the most abundant free amino acid in tea leaves, and the main source of a liquor's savory freshness and sweetness. Think of it as the leaf's built-in MSG. In "high heat destroys theanine," the word "destroys" covers two different things:

Claim A: heat decomposed it High temperature broke the molecule apart, and the absolute amount in the cup really fell. This is answered by measuring concentration.

Claim B: it is still there, just drowned out It dissolves out as usual, but catechins and caffeine dissolve out faster and in greater amounts, pushing down the share of sweetness and savoriness. This is answered by measuring proportion.

The two taste the same: either way the cup is less sweet and more bitter. Only instruments can tell them apart.

When you see a figure for "how much it dropped," check first whether it measured concentration or proportion. So far no study supports Claim A; the data for Claim B are consistent enough to chart.

2. Where "boiling water cuts theanine by 25%" comes from

The number comes from a 2025 Longjing study in _Food Chemistry: X_: four water temperatures, 4 minutes each, a panel of nine trained tasters, and the paper does report a 25.20% decline [2]. The figure is often screenshotted and shared on its own, and few people look at the y-axis.

That 25% is a proportion. Concentration asks how many milligrams of theanine are in the cup; proportion asks what share theanine makes up of everything measured. When the denominator grows, the proportion falls, even if the numerator has not changed at all. Only concentration can answer Claim A, and this paper reports proportion: at 100 degrees more material dissolves out, so theanine's share naturally gets smaller.

The figure caption says "contribution proportion," and the y-axis is labeled `Proportion (%)`. The methods section explains that each compound's number is its own signal divided by the summed signal of every compound detected in the cup, and theanine accounts for only 2.6% to 3.5% of that total [2].

〔圖〕Schematic of the difference between proportion and concentration: the numerator was never measured, while the denominator grows because catechins and caffeine dissolve out in quantity, so the proportion falls

This study did not measure theanine concentration. Boxes are schematic, not to scale; figures come from its supplementary data. The study covers Longjing only (a tender-plucked green tea), and its conclusions cannot be carried straight over to oolong or pu-erh [2].

In the same paper, concentration goes up

The authors also used an older colorimetric assay to measure directly how many milligrams of free amino acids, all added together, were in the liquor, and for both grades of Longjing that number rose with water temperature. The conclusion states it plainly: the 70-degree cup scored highest for umami, but its free amino acid concentration was significantly lower than the 100-degree cup's; what fell with temperature was the proportion [2].

The extra bitterness is real: premium Longjing's umami-mellowness score fell from 8.10 to 5.60, and the standard grade's bitterness and astringency rose significantly. The authors attribute this to differential extraction: hotter water also pulls out the larger flavonol glycosides and theaflavins, diluting umami's share, rather than heat breaking theanine apart [2].

3. Six Taiwanese teas: amino acids rise too, just more slowly

In 2012 Taiwan's Tea Research and Extension Station (TRES) ran a much broader experiment: six Taiwanese teas, each brewed at 80°C, 90°C, and boiling (3 grams to 150 ml). The paper says it directly: the concentrations of total catechins, caffeine, and total free amino acids "all increased with brewing temperature." Of the eighteen sets of figures, only black tea's amino acids were slightly lower at boiling than at 90°C, and its three temperatures fall in the same statistical group; no set showed a significant decline [1].

The raw numbers: six teas × three water temperatures, as measured

Tea ｜ Total catechins ｜ Caffeine ｜ Free amino acids

Green tea (Biluochun) ｜ 436ᶜ → 522ᵇ → 728ᵃ ｜ 188ᶜ → 287ᵇ → 397ᵃ ｜ 250ᵇ → 264ᵇ → 312ᵃ

Wenshan Baozhong ｜ 282ᶜ → 343ᵇ → 469ᵃ ｜ 103ᶜ → 179ᵇ → 286ᵃ ｜ 177ᵇ → 212ᵃᵇ → 235ᵃ

High-mountain tea ｜ 195ᶜ → 282ᵇ → 487ᵃ ｜ 72ᶜ → 135ᵇ → 267ᵃ ｜ 146ᵇ → 175ᵇ → 220ᵃ

Dong Ding oolong ｜ 264ᶜ → 364ᵇ → 554ᵃ ｜ 107ᶜ → 175ᵇ → 290ᵃ ｜ 120ᵇ → 127ᵇ → 182ᵃ

Oriental Beauty ｜ 78ᶜ → 111ᵇ → 140ᵃ ｜ 149ᶜ → 281ᵇ → 432ᵃ ｜ 68ᵇ → 78ᵇ → 106ᵃ

Black tea ｜ 39ᶜ → 49ᵇ → 63ᵃ ｜ 215ᶜ → 307ᵇ → 417ᵃ ｜ 205ᵃ → 230ᵃ → 221ᵃ

Units are mg/L; each cell reads 80℃ → 90℃ → boiling. Superscript letters are statistical groupings; cells sharing a letter did not differ significantly. Re-typeset by this article; data from Table 8 of the Taiwan Tea Research Bulletin 31: 99–114 (2012) [1]. Free amino acids were measured with a colorimetric total assay, which gives the sum of all free amino acids; theanine was not measured on its own.

All three rise, at very different speeds

Converted to percentage increases, the order is the same for all six teas: caffeine rises most, catechins next, free amino acids least. All percentages and ratios in this article are our own calculations from the paper's figures.

Increases calculated by this article from Table 8 [1]. Black tea's free amino acids fall in one statistical group across all three temperatures, with no significant difference, so no increase is shown.

The absolute amount of amino acids did not fall; their share did. Divide amino acids by caffeine as a rough "sweet-savory versus bitter" ratio, and all six teas drop to 40–60% of their original value:

Ratios calculated by this article from Table 8 [1]; a rough proportion only, not an official quality metric.

Black tea shows it most clearly: free amino acids go 205 → 230 → 221, with no statistical difference across the three temperatures, while caffeine rises from 215 to 417, nearly doubling [1]. These data support Claim B; Claim A does not hold.

There are also studies that measure theanine on its own. A 2025 study in _Foods_ covered 80 and 100 degrees, six oxidation levels, and three steeping times, quantifying theanine individually by instrument, and found that the differences caused by temperature were not statistically significant [3]. It tested only two to six cups per tea, so failing to find a difference does not mean there is none; it is best read alongside the two studies above.

At brewing temperatures and timescales, there is no evidence that heat breaks theanine down, and the evidence in the other direction is consistent. No study has yet combined individual theanine quantification, a continuous 70-to-100-degree gradient, and multiple tea types, so it cannot yet be called proven impossible.

4. What heat does to catechins: mostly, it changes their shape

The rumor often comes with a second one: high heat destroys catechins. Catechins are the main source of astringency in tea; think of them as the leaf's natural sunscreen. Heat does break them down. What matters is how much breaks down in the three to five minutes it takes to brew a cup.

A 2008 study in the _Journal of Agricultural and Food Chemistry_ heated EGCG solutions from 25 to 165 degrees and built a reaction-rate model: at a sustained boil for thirty minutes, the predicted EGCG loss is 12.4%, consistent with the 10% to 15% measured in an independent paper [5]. That is thirty minutes of continuous boiling; a cup of tea usually steeps for three.

Green: EGCG thermal degradation (the 30-minute figure is the paper's prediction; the 5-minute figure is this article's estimate from its published parameters) [5]. Gold: the range of extra total catechins extracted when moving from 80℃ to boiling water (this article, from Table 8) [1]. The baselines differ; they are shown together only to compare magnitudes.

Using the same paper's parameters, thermal degradation over a five-minute boiling-water steep comes to about 1.25%, and about 0.6% at eighty degrees. This is a model of pure EGCG solution; real tea liquor may degrade faster, and it covers only the EGCG–GCG pair, not all catechins [5]. The same move from eighty degrees to boiling extracts 62% to 150% more total catechins [1]. Extraction is two orders of magnitude larger than degradation.

The 44 and 98 degrees flagged in that study are often reported as "98 degrees is a critical point." The paper says the individual reactions have no turning point; those two temperatures are where three rate curves overtake one another in rank. Above 98 degrees, the isomerization of EGCG into GCG outpaces its degradation [5]. The molecule changes shape and is still a catechin.

Another study, which heated eight pure catechin solutions, kept exact accounts: after half an hour at ninety degrees, only 73.4% of the epicatechin remained in its original form, and the missing 25.7% had all turned into its mirror-image isomer, the same atoms flipped over, for a total of 99.1% [15]. It has been measured in real tea liquor as well: steeped for up to five hours, the 95-degree batch had significantly higher antioxidant capacity than the 60-degree batch at every time point [6]. Both studies have limits: the first used pure water, went no higher than 90 degrees, and its shortest interval was half an hour; the second had only two temperatures, 60 and 95, and a single green tea sample [15][6].

Temperature and time on one table

A 2011 German study laid out both variables together. Water from 70 to 100 degrees was poured onto one green tea and then left to cool naturally (the authors note this is how people normally brew), with an extra group kept at a sustained boil; the four main catechins were measured at 3, 5, and 7 minutes. Comparing the extremes, 70 degrees for 3 minutes versus 100 degrees boiled for 7 minutes, the latter delivered more than three times the EGCG. The study used a single green tea at a 1:100 tea-to-water ratio, did not measure theanine, and reported no significance tests between temperatures [14].

Time matters as much as temperature: at the same seventy degrees, going from 3 to 7 minutes added seventy percent more EGCG [14]. The theanine-only study above found the same: temperature differences were significant only for oolong, while time significantly raised both compounds [3].

Of the four catechins, only EGC goes the other way, dropping somewhat under a sustained boil (the authors judge it the only one undergoing oxidative degradation), yet seven minutes at a sustained boil still gives 1.8 times the EGC of three minutes at seventy degrees [14].

To adjust bitterness and astringency, adjust steeping time first.

5. The same hundred degrees: a three-minute steep and a twelve-hour roast

Theanine does break down during roasting, by two routes. In one, it loses its amino group on its own and splits into aroma molecules via intermediates, with no sugar needed. In the other, it meets reducing sugars in the leaf and goes through the Maillard reaction, which speeds up the first route and produces more aroma [7]. Part of hojicha's toasty aroma and a Wuyi rock tea's roasted character comes from theanine consumed this way.

The 2022 study summarizes industry roasting conditions: Japanese hojicha is usually roasted at 160 to 200 degrees for 15 to 30 minutes, while Wuyi rock tea goes low and slow, with a "light fire" at 100 to 120 degrees for 12 to 15 hours [7].

A light roast runs at almost the same temperature as the boiling water in your kettle, so the reason brewing does not decompose theanine while roasting does is not temperature. It comes down to three other things.

〔圖〕Brewing versus roasting, side by side: the temperatures overlap; what differs is duration, wet versus dry state, and whether catechins are present to suppress the reactions

Roasting conditions from the study's summary of industry practice [7]; this figure shows differences in conditions only and implies no decomposition ratio.

First, time: a few minutes against a dozen or more hours, a difference of one hundred to three hundred times. Second, moisture: to rule out reactions involving water, that study dried the leaves to 0.15% moisture before heating, whereas brewing happens entirely in water. Third, the surrounding compounds: the researchers found that catechins strongly suppress the formation and release of theanine's degradation products, and tea liquor is exactly a catechin-rich environment. As for the temperature at which decomposition starts, there is no clean answer: the same paper detected trace products after one hour at 50 degrees, while after an hour at 160 degrees with added glucose, the 56 compounds identified still added up to only trace amounts [7].

What this study could not measure Theanine is certainly consumed, but no one knows how much. The study measured detected volatile products and never weighed how much theanine remained (on inspection, its supplementary data is the pre-heating input table; no mass balance was done anywhere in the paper), and non-volatile products were not measured. In real tea leaves, those products were suppressed by other tea constituents to the point of "greatly inhibited, trace-level, or even undetectable," so consumption cannot be back-calculated from product yields either [7]. For that reason this article gives no figure for how much roasting decomposes. For more on roasting, see Catechins and the Roast.

6. What temperature should your tea get

TRES's guideline is that water temperature depends on the tea: "the more tender the pluck, the lower the brewing temperature" [1]. Besides measuring compounds, the same 2012 study brewed and tasted all six teas one by one, with these results: Tea ｜ Suggested starting temperature ｜ What boiling water does ｜ Practical note

Green tea (Biluochun) ｜ 80–90℃ ｜ Not as good as 80–90℃ ｜ The most tender pluck of the six, right at the "tender pluck, cooler water" end

Wenshan Baozhong ｜ 90℃–boiling ｜ Within the best range ｜ It was the 80℃ cup that rated lower

High-mountain tea ｜ Boiling ｜ Best: rich aroma, sweet and smooth ｜ At 80℃ the leaves "could not unfurl"; if the leaf will not open, nothing else matters

Dong Ding oolong ｜ Boiling ｜ Best ｜ The paper specifically notes it is a "high-temperature roasted" tea

Oriental Beauty ｜ 80–90℃ ｜ "Slightly sour, stuffy and astringent" ｜ The 80℃ cup was described as "honeyed aroma, sweet taste"

Black tea ｜ 90℃–boiling ｜ Within the best range ｜ 80℃ was the lowest-rated of the three temperatures

Compiled from the Taiwan Tea Research Bulletin 31: 99–114 (2012) [1]; the original table is not reproduced. The table shows TRES's tasting conclusions; this article gives no separate brewing advice.

Of the six teas, only the green tea and Oriental Beauty did better below a boil. Applied as a general rule, "never use 100 degrees" gets it backwards for high-mountain tea and Dong Ding oolong, the teas Taiwanese drinkers reach for most. Brewing green tea and Oriental Beauty cooler slows down the bitter and astringent compounds; the sweetness and savoriness do not decrease [1][2].

This table lists only the six teas with actual tasting data, so it is shorter than the usual brewing-temperature chart, but these six are exactly what Taiwan drinks every day. TRES gave only these descriptions; it drew no general rule about which kinds of tea want high heat, and offered no mechanism. For other teas, and for the practical side (exact temperatures, what to do without a thermometer), see the water temperature guide.

What different countries say: seven different tables

Four of them teach people how to brew a good cup. The first two come from government bodies, the other two from trade associations:

Japan MAFFGovernment ministry · everyday drinking

Gyokuro 50–60℃ (with 10 ml per cup, 150 seconds); premium sencha 70℃; mid-grade sencha 90℃; bancha, genmaicha, and hojicha take "boiling water" [18]

Taiwan TRESGovernment research station · everyday drinking (descriptive tasting)

Six Taiwanese teas; see the table above [1]

UK Tea & Infusions AssociationTrade association (non-governmental) · everyday drinking

Black tea 90–98℃ [20]

Canada's THACTrade association (non-governmental) · everyday drinking

Black tea 100℃ for 3–5 minutes; white, green, and oolong tea all listed together at 80℃ [21]

Three more also specify a water temperature, for sensory evaluation, compositional analysis, and food safety respectively:

ISO 3103:2019International standard · sensory evaluation

Boiling water specified for both black and green tea [4]

GB/T 23193—2017Chinese national standard · compositional assay

Boiling water, 100℃ water bath, 30 minutes [16]

Germany's BfRGovernment risk-assessment agency · food safety

Herbal infusions must be made with boiling water; the reason is microbial risk, not flavor [19]

⚠️ All seven are recommendations, not experiments; they show only how various authorities teach people to brew. The institution types and purposes are this article's own annotations.

The one point all seven agree on is boil the water first. MAFF's third principle reads "湯は必ず沸騰させてから使用する" (always bring the water to a boil before use), and the Canadian guide likewise says to bring it to a rolling boil and then let it cool [18][21]. Japan's official guidance tells you to always boil the water, then let it come down to what that tea needs; the low temperatures refer to the moment of pouring.

The reason the manual gives for gyokuro's low temperature is to let the theanine dissolve out fully while holding back bitterness and astringency [18]. Nowhere in the document does it say that high heat destroys theanine.

Where they disagree: for oolong, the Canadian association gives 80℃, while TRES found high-mountain tea and Dong Ding oolong better at a boil, a 20-degree gap [1][21]; the Canadian table also puts white, green, and oolong tea in a single row. The Japanese manual does not cover black tea at all, though "black tea, 90–100℃" is often attributed to MAFF [18].

Evaluation and testing use boiling water

In the specification table of ISO 3103:2019, the international protocol for sensory evaluation, the water-temperature column for both black and green tea reads Boiling (approx. 100 °C) [4]. China's national standard GB/T 23193—2017 specifies that to measure a tea's theanine, 1 gram of ground sample is combined with 100 ml of boiling distilled water and extracted in a 100-degree water bath for 30 minutes. This condition was introduced in the 2017 revision; the previous edition used an 80-degree bath for 45 minutes [16].

ISO 3103 is an evaluation protocol: 2 grams to 100 ml, much stronger than everyday brewing, intended to magnify differences between samples. GB/T 23193 is a quantification method, not a drinking method: the sample is ground first, the liquid is more dilute than a brewed cup, and it is not telling anyone to steep tea for thirty minutes. Analytical methods can tolerate small, consistent losses, so the standard shows only that thirty minutes in boiling water causes no analytically meaningful loss, which is not the same as zero loss; that revision also rewrote the whole principle and sample preparation, so water temperature was not the only change [16].

Further reading: [caffeine in cold-brew tea](/articles?slug=cold-brew-caffeine), [water for tea](/articles?slug=waterfortea) (more important than temperature), and [why high-mountain tea tastes sweeter](/articles?slug=high-mountain-theanine).

7. "Hard on the stomach" and "destroys the vitamins"

These two claims often travel with the rumor, and no study has yet directly compared boiling water with cooler brewing for either.

The stomach. No study has compared the effects of brewing at 100 degrees and at 80 degrees on the stomach. The closest is a 2019 meta-analysis that pooled 23 studies into 30 estimates, comparing tea drinkers with non-drinkers for gastroesophageal reflux: taken together, no clear association, with wide disagreement between studies, and the three highest-grade prospective studies found no association. The authors concluded there was no significant overall association, but that the East Asian subgroup might show an increase, and so suggested that people who already have reflux symptoms watch their intake [9].

In its limitations section the paper states that tea type, oxidation level, polyphenol concentration, and the temperature of the tea are all confounders it could not handle [9]. So it can neither prove nor refute "hot-brewed tea harms the stomach"; epidemiology has not yet asked that question. What can be offered is dose: a 150 ml cup of high-mountain tea brewed with boiling water contains about 40 mg of caffeine, versus about 11 mg at eighty degrees [1]; a cup of Americano has 100 to 150 mg.

IARC's 65 degrees is about drinking temperature The International Agency for Research on Cancer classifies "drinking very hot beverages above 65℃" as Group 2A (probably carcinogenic to humans), in relation to esophageal squamous cell carcinoma; in the same assessment, coffee itself and mate that is not very hot were both placed in Group 3 [10]. ⚠️ This classification concerns the temperature at which a drink is consumed, not the brewing water, and the monograph itself keeps the two apart: in the section defining "very hot beverages," it cites the ISO requirement that tea samples for sensory testing "should be brewed with boiling water" [10]. The monograph's figures: these hot drinks are typically 71 to 85 degrees when served and 50 to 70 degrees when drunk. 65 degrees falls in the middle, and whether you go over it depends on how long you wait. In the study from Iran's high-incidence esophageal cancer region, the exposure variable was defined as "how many minutes one usually waits after pouring boiling water over the tea before drinking" [10]: everyone used boiling water; the difference was whether they let it cool. Group 2A is a formal classification, with limits in its evidence grade and exposure assessment [10]. This section only explains what the classification covers; it is not health advice.

The vitamins. Measured vitamin C in tea liquor is 1.43 to 2.45 mg per 100 ml; a kiwifruit has about 70 mg. Tea was never a source of vitamins to begin with [8].

That study used 7 grams of tea per liter, so converting to everyday strength gives only an order-of-magnitude estimate; its highest temperature was 85 degrees, with no boiling water and no 80-versus-100 comparison, so it cannot say whether 100 degrees destroys vitamins [8].

8. A thousand years of watching the water: the ancients were concerned with something else

"Don't let the water boil too long" goes back twelve hundred years in Chinese tea writing and is often cited in support of the rumor. In the boiling chapter of Lu Yu's _Classic of Tea_ (茶經), after the third boil, he writes "beyond this the water is old and not fit to use" [11]; Cai Xiang's _Record of Tea_ (茶錄) opens its section on judging the water with the words "judging the water is hardest of all" [12]; Emperor Huizong's _Treatise on Tea_ (大觀茶論) is more practical: if the water has gone old, "add a little fresh water to it" [13].

Cai Xiang explains the difficulty himself: "sunk in the ewer as it boils, it cannot be discerned"; the ewer was opaque, so you could not see how far the boil had gone [12]. Tang tea was boiled and Song tea was whisked, and "old water" meant water still on the fire, heating continuously, which is different from today's water, boiled and then poured into the pot [11][12].

"Prolonged boiling lowers the oxygen content," "old water's molecular clusters grow larger," and "repeated boiling produces nitrites" are later explanations that do not appear in the classical texts, and this fact-check found no academic source supporting them [17].

The takeaway, in one cup

One

"A hundred degrees destroys theanine" does not hold up. At brewing temperatures and timescales, no study has measured a drop in theanine's absolute concentration, and the evidence in the other direction is consistent: in six Taiwanese teas and two grades of Longjing, wherever concentration was measured, hotter water extracted more [1][2]; the study that quantified theanine on its own found no significant difference due to temperature [3]. The "25% drop" refers to a proportion.

Two

"Green tea tastes better brewed cooler" is true, and the reason is a flavor trade-off: the sweetness still dissolves out, but it is masked by bitterness and astringency that dissolve faster [1][2]. Only two of the six teas should be brewed cooler; high-mountain tea and Dong Ding oolong are best at a boil [1].

Three

The temperature with health evidence behind it is the sip: IARC's "above 65 degrees" refers to drinking temperature, and whether you go over it depends on how long you wait [10].

References

1. 戴佳如、劉天麟、陳國任、林金池（2012）。〈水質及水溫對茶湯品質及化學成分之影響〉。《臺灣茶業研究彙報》31: 99–114，行政院農業委員會茶業改良場（今農業部茶及飲料作物改良場）。科普摘要頁：https://www.tbrs.gov.tw/ws.php?id=5283
2. Deng S, Cao Q-Q, Gao Y, Wu W, Chen J-X, Wang F, Zou Q, Xu F, Cao X, Sun W, Yin J-F, Xu Y-Q (2025). Elucidating the effect of brewing temperature on the sensory quality of Longjing tea based on multi-scale molecular sensory science. _Food Chemistry: X_ 28:102635. DOI: 10.1016/j.fochx.2025.102635
3. Ayakdaş G, Ağagündüz D (2025). Determination of L-Theanine and Caffeine Contents in Tea Infusions with Different Fermentation Degrees and Brewing Conditions Using the Chromatographic Method. _Foods_ 14(13):2313. DOI: 10.3390/foods14132313
4. ISO 3103:2019 _Tea — Preparation of liquor for use in sensory tests_（第二版，取代 ISO 3103:1980）。ISO/TC 34/SC 8。標準頁：https://www.iso.org/standard/73224.html
5. Wang R, Zhou W, Jiang X (2008). Reaction Kinetics of Degradation and Epimerization of Epigallocatechin Gallate (EGCG) in Aqueous System over a Wide Temperature Range. _J Agric Food Chem_ 56(8):2694–2701. DOI: 10.1021/jf0730338
6. Jin Y, Zhao J, Kim EM, Kim KH, Kang S, Lee H, Lee J (2019). Comprehensive Investigation of the Effects of Brewing Conditions in Sample Preparation of Green Tea Infusions. _Molecules_ 24(9):1735. DOI: 10.3390/molecules24091735
7. Li M, Ho C-T, Wang J, Hu Y, Zhai X, Zhang L, Wan X, Yang X (2022). Formation of Volatile Heterocyclic Compounds and Open-Chain Amides of Theanine in Model Systems with Glucose, Tea Leaves, and Tea Extract under Tea-Roasting Conditions. _J Agric Food Chem_ 70(22):6737–6746. DOI: 10.1021/acs.jafc.2c02039
8. Oracz J, Królak K, Kordialik-Bogacka E, Żyżelewicz D (2025). Optimizing brewing conditions for low-temperature green tea infusions: Insights into functional and nutritional properties. _Food Chemistry_ 474:143241. DOI: 10.1016/j.foodchem.2025.143241
9. Cao H, Huang X, Zhi X, Han C, Li L, Li Y (2019). Association between tea consumption and gastroesophageal reflux disease: A meta-analysis. _Medicine (Baltimore)_ 98(4):e14173. DOI: 10.1097/MD.0000000000014173
10. IARC Monographs on the Identification of Carcinogenic Hazards to Humans, Volume 116: _Drinking Coffee, Mate, and Very Hot Beverages_（2018）。國際癌症研究總署（WHO）。NCBI Bookshelf：https://www.ncbi.nlm.nih.gov/books/NBK543949/
11. 陸羽《茶經・五之煮》（唐，約 760–780 年成書）。中國哲學書電子化計劃：https://ctext.org/wiki.pl?if=gb&chapter=158743
12. 蔡襄《茶錄》〈候湯〉〈熁盞〉（北宋，1051 年進呈）。維基文庫：https://zh.wikisource.org/zh-hant/茶錄
13. 宋徽宗趙佶《大觀茶論・水》（北宋，1107 年）。維基文庫：https://zh.wikisource.org/zh-hant/大觀茶論
14. Zimmermann BF, Gleichenhagen M (2011). The effect of ascorbic acid, citric acid and low pH on the extraction of green tea: How to get most out of it. _Food Chemistry_ 124(4):1543–1548. DOI: 10.1016/j.foodchem.2010.08.009
15. Fan F-Y, Shi M, Nie Y, Zhao Y, Ye J-H, Liang Y-R (2016). Differential behaviors of tea catechins under thermal processing: Formation of non-enzymatic oligomers. _Food Chemistry_ 196:347–354. DOI: 10.1016/j.foodchem.2015.09.056
16. GB/T 23193—2017《茶葉中茶氨酸的測定 高效液相色譜法》（代替 GB/T 23193—2008）。中華人民共和國國家質量監督檢驗檢疫總局、中國國家標準化管理委員會發布（2017-11-01 發布／2018-05-01 實施），全國茶葉標準化技術委員會（SAC/TC 339）歸口
17. 本文不採用的說法：「水煮久了含氧量下降」「老水的分子團變大」「反覆煮沸產生亞硝酸鹽」，古籍無據，也查無學術研究支持。另有一份日本給茶機成果表常被引用來證明「94 度破壞茶胺酸」，但表中三組數據的品種、茶葉量與攪拌時間都不同，無法比較溫度，本文同樣不採用。
18. 農林水産省《茶の淹れ方マニュアル》（英文版 _The proper way to brew Nihoncha_）。日文版 PDF：https://www.maff.go.jp/j/seisan/tokusan/cha/attach/pdf/ocha-3.pdf ；官方英語版：https://www.maff.go.jp/j/seisan/tokusan/cha/attach/pdf/ocha-5.pdf ；母頁：https://www.maff.go.jp/j/seisan/tokusan/cha/ocha.html （查閱 2026-08-14）
19. Bundesinstitut für Risikobewertung (BfR)（德國聯邦風險評估研究所）：《Temperierte Heißwasserspender für Kräuterteeaufgüsse nicht geeignet》，Stellungnahme Nr. 045/2005（2005-09-13）；併同新聞稿 Nr. 37/2005（2005-12-14）。https://www.bfr.bund.de/cm/343/temperierte_heisswasserspender_fuer_kraeuterteeaufguesse_nicht_geeignet.pdf （查閱 2026-08-14）
20. UK Tea & Infusions Association (UKTIA), "The Perfect Brew". https://www.tea.co.uk/make-a-perfect-brew （查閱 2026-08-14，頁面標示 © 2026）
21. Tea and Herbal Association of Canada (THAC), "How to Brew". https://www.tea.ca/learn/how-to-brew/ （查閱 2026-08-14，頁面標示 © 2026）