Kukicha & Karigane: The Twig Teas

Kukicha is Japanese green tea made from the stems and leafstalks that a sorting machine pulls out of the leaf during refining. Japan’s labelling standard defines it by that machine and treats kukicha and bōcha as one name. Karigane, the premium word, appears in no standard at all. Stems carry three to nine times the theanine of the leaf beside them and less caffeine, which is why the cup is sweet. Brew ordinary kukicha near boiling.

The Green Tea Labelling Standard of the Japan Tea Central Public Interest Association names seventeen kinds of tea. Sixteen of them describe a leaf, or an operation performed on one. Entry eleven describes a machine.

茎茶又は棒茶 — kukicha (茎茶), or bōcha (棒茶): “tea containing many of the stems and leafstalks of tea, sorted out by a stem separator or the like during the refining process of aracha” ¹.

The machine is the mokkei bunriki (木茎分離機), the wood-and-stem separator. It sits near the end of the line, after the crude tea called aracha (荒茶) comes in from the farm and before the finished bag goes out, and its job is to pull the twigs out of the needles. Everything else about this tea follows from what that machine is for.

What the separator takes out

It takes out stems and yōhei (葉柄), the leafstalks — the short stalk joining a leaf to the shoot. Ippodo describes the raw situation plainly: a sencha leaf arrives from the field still clinging to the soft stem it sprouted from, and processing splits the two apart ¹¹.

Read entry eleven once more and notice the hedge. Tea containing many stems and leafstalks ¹. The standard declines to say a bag of kukicha is all stem, and the Global Japanese Tea Association agrees: kukicha is mostly stems, with some leaf still in the mix ⁶.

Kukicha has two siblings in the standard, and the three of them are defined the same way. Entry nine, konacha (粉茶): the powder sifted out in refining. Entry ten, mecha (芽茶): the bud tips sifted out in refining sencha or gyokuro ¹. Sencha is what the standard says you made. These three are what fell off the sieve while you made it. The Japanese trade calls them demono (出物), the sortings, and files bancha alongside them.

Now the part that decides everything downstream. The GJTA gives the reason the stems come out: to make the leaf teas uniform and better-looking ⁶. A tight dark needle is what a buyer of sencha is paying to see, and a pale twig lying across the pile spoils the look of it. So the separator is, in its origins, a cosmetic machine. It removes a part of the shoot on the grounds that the part is the wrong shape.

Set that beside the tea room, where a bowl is admired for the very irregularity a factory would reject, and a break repaired in gold makes the piece more valuable than it was whole. The refining line runs on the opposite instinct. It sorts for sameness.

Karigane, shiraore, bōcha

Here is where most write-ups go soft, so take the sources one at a time.

ITO EN, whose tea encyclopedia is about as mainstream as Japanese tea reference gets, defines kukicha as the stems of new shoots extracted by a sorting machine during the refining of gyokuro or sencha, and says that the premium ones — off high-grade gyokuro or sencha — “are called karigane and prized” ⁴. Kodansha’s beverage dictionary defines karigane (雁ヶ音, “the cry of the wild goose”) as premium kukicha, using stems separated during the production of “gyokuro and the like” ⁵. That hedge is the dictionary’s, not mine.

English sources push further. Karigane historically meant kukicha made from gyokuro leaf, and the term “has become diverse, and is also used for any kukicha of high quality” ⁷⁸.

Then the map. In Kyushu the same tea is called shiraore (白折) — Kyoto says karigane, Kyushu says shiraore, and the tea is the same tea ⁹. Wikipedia states the rule from the other end: karigane specifically from Kyushu is shiraore ⁸. Both names cover stem tea made from high-grade sencha and from gyokuro alike ⁹, so a Kyushu shiraore may be either, and the regions map the districts that argue about it.

So: the strictest sense of karigane is gyokuro stems. The working sense, the one ITO EN and the dictionary and the Kyushu trade use, is stems off gyokuro or high-grade sencha. The loosest sense is any good kukicha. All three are in live use, and the honest answer to “which is correct” is that the usage varies and always has.

There is no standard anywhere in Japan that defines karigane. I extracted the full text of the labelling standard to check: 雁ヶ音 appears nowhere in it, nor does 白折, and the supplementary table of permitted names outside the main list runs only to shincha, kocha and the numbered harvests ¹. The committee that regulates whether a bag may call itself 極上 has never troubled itself with karigane, because karigane is not a grade. It is a compliment.

ITO EN’s distinction is the one that matters and the standard is blind to it. Tender stems off a hand-picked first flush and coarse machine-cut sticks both satisfy entry eleven. The market’s premium word is unregulated, and the market’s real quality signal has no official name at all.

Why the stem is the sweet part

The chemistry here is better than the folklore, and it has been measured.

Theanine, the amino acid that carries sweetness and umami, is made in the tea plant’s roots, mostly in late autumn and winter. In spring it moves up into the new shoots through the xylem ². What happens when it arrives is the interesting part. A 2025 paper in Plant Physiology puts it in one line:

Theanine … is primarily synthesized in roots of tea plants and transported to new shoots, where it is mainly distributed to the young stem; however, tea is predominantly produced from young leaves.

The stem is where the theanine goes, and the leaf is what Japan sells. Measured across eleven cultivars, the theanine ratio between the first stem and the first leaf of a new shoot ran from 3.3 to 8.8 times in the stem’s favour ². A transporter protein called CsCAT1, sitting in the stem’s vascular tissue, does the holding: express it strongly and theanine stays in the stem, knock it down and more of it reaches the leaf ².

Two caveats, because this is the claim vendors inflate. Those measurements are of the fresh shoot, not of the bag on your shelf, and a bag holds thicker stems than the paper’s first stem. The numbers explain the direction of the flavour. They do not survive being restated as a fact about your cup.

The other half of the trade runs the opposite way. Comparing stems and leaves of matched tenderness in a Chinese green tea, caffeine, total polyphenols and total catechins all came out significantly higher in the leaves, while total free amino acids came out higher in the stems ³. That study is Lu’an Guapian from Anhui rather than yabukita from Uji, so read the direction and leave the decimals alone.

Put the two together and kukicha’s whole character falls out. The sweet compound concentrates in the stem. The bitter and stimulating ones concentrate in the leaf. The separator divides the shoot along almost exactly that line, and then the industry prices the sweet half at a discount because it is the wrong shape.

Shading is a second, separate lever on the same chemistry — sunlight converts theanine into catechins, so twenty days under cover leave the leaf loaded with it, which is the whole argument of gyokuro. A gyokuro karigane gets both effects at once: shaded plant, and the stem of it.

The number nobody has

Japan measures its teas. The national Standard Tables of Food Composition give caffeine and tannin per 100 grams of infusion for gyokuro, sencha, kamairicha, bancha, hōjicha and genmaicha, each on a stated reference brew.

They carry no entry for kukicha. None for karigane, bōcha, mecha or konacha either ¹². The three teas the separator makes are absent from the country’s own composition tables.

So the honest position on kukicha’s caffeine is that it is low, well-supported in direction, and unmeasured by anyone with authority. A Japanese tea seller writing a whole page on the question ends up saying that hōjicha runs about 20 mg per 100 g and kukicha is presumed lower ¹⁰ — 推測されます, the verb doing the work. When a shop with stock to move declines to give you a number, believe the absence. If you want a figure you can lean on, the measured ones sit in matcha, caffeine and L-theanine.

How to brew it

The sources split, and the split is instructive.

Three guides say 70–80°C. The Kyoto house that has been refining this material since 1717 says boiling ¹¹.

Follow Ippodo. The 70–80°C rule belongs to sencha, where cool water exists for one purpose: to leave the bitter catechins in the leaf while the sweet theanine comes out. Run that reasoning on a stem and it collapses, because the catechins that the rule protects you from largely stayed behind in the leaf the separator threw this twig out of. The GJTA is blunt about the consequence — boiling water for several minutes will not make kukicha too bitter ⁶. Cool water on ordinary kukicha guards against a bitterness that was never in the pot, and costs you the aroma that heat would have lifted. It is sencha’s discipline, applied by reflex to a twig.

One exception, and it is the expensive one. Brew a gyokuro karigane at 70°C or a little below ⁹. You bought that tea for thickness and sweetness ahead of fragrance, and cool water brings the sweetness forward. Ordinary kukicha wants a kettle. Gyokuro karigane wants the patience you would give the leaf it came off.

Both re-steep, and neither goes far. Flavour drops off sharply by the second infusion ⁷, which is what you would expect of a material that gives everything up in thirty seconds. Every other Japanese tea and its numbers are in the temperature and time guide.

The roasted branch

Roast kukicha and it becomes something else. Stem-based hōjicha brews lighter and sweeter than the leaf-based kind and sits lowest of all in caffeine, and the famous one is kaga bōcha (加賀棒茶) from Kaga in Ishikawa. That tea, its 1902 origin and the chemistry of what a roaster does to a leaf belong to hōjicha, which is built on the question. Note only where the boundary runs: kukicha is the raw material, and the roast is a separate operation the standard defines separately ¹.

What to buy

Ignore the word karigane. It is unregulated, it means three different things depending on who printed it, and it is the one word on the bag nobody can be held to.

Read the base instead. “Gyokuro kukicha” and “sencha stems” are claims about raw material that a seller can be challenged on. Ippodo prints exactly that and never reaches for the romance: its stem tea is sold as Stems Sencha, “a type of kukicha”, off the same highest-grade leaves as its sencha ¹¹. A house with three centuries of reputation would have every commercial reason to write karigane on the packet. It writes the leaf.

Then look at the stems. ITO EN’s tell is that the glossier and brighter green they are, the sweeter the tea, while thick reddish-brown sticks mark the machine-harvested end of the category ⁴. That is the distinction the standard collapsed into one entry, and you can see it through the bag before you pay.

Which brings me to what I think this tea is. A gyokuro karigane is the stem of the most expensive leaf in Japan, off the same shaded bush, cut on the same day, holding several times the theanine that the leaf beside it held — sold cheaper because it came out of the sieve rather than the sencha. The trade files it under demono, the stuff that fell out. The plant files it under where the sugar went.

The separator has been sitting in entry eleven since April 1991, through every revision to July 2026 ¹. It was installed to make the leaf look like leaf. What it has been quietly doing for a century is dividing each shoot into the part Japan sells and the part Japan drinks — and among the seven teas, it is the only one whose existence is an accident of tidiness.