Evening, a cup on the table, and the hand stops: is this a reasonable hour to be drinking it? Next morning brings a latte in a paper cup and the thought that it might be made at home. An old bag of sencha surfaces at the back of the kitchen, and the question is whether it could be roasted.

The three questions at the entrance

Those three sit at the entrance to hojicha, more or less on their own. The answers are in chapters 1 and 2; what goes here is only where each one gets decided.

Caffeine turned out to be a question of conditions

Caffeine turned out to be a question of conditions. The ministry's table uses different infusion conditions for each kind of tea (Figure 0-2-1). Gyokuro is given as 160 mg per 100 mL of infusion, eight times hojicha's 20 mg. Gyokuro's conditions, though, are 10 g of leaf to 60 mL of water, or 16.7 g per 100 mL of water, more than seven times hojicha's 2.31 g. Lined up again per gram of leaf, the eightfold gap shrinks to about 1.1 times [a calculated value]. Putting the conditions back on one footing is chapter 1.

Figure 0-2-1Caffeine and infusion conditions by kind of tea

TeaCaffeine (mg per 100 mL of infusion)Infusion conditions (leaf, water temperature, volume, time)Leaf per 100 mL of water (calculated)
Gyokuro16010 g leaf・60°C・60 mL・2.5 min16.7 g
Sencha2010 g leaf・90°C・430 mL・1 min2.33 g
Hojicha2015 g leaf・90°C・650 mL・0.5 min2.31 g
Genmaicha1015 g leaf・90°C・650 mL・0.5 min2.31 g
Drawn from "About caffeine", Ministry of Agriculture, Forestry and Fisheries. Values from the Standard Tables of Food Composition in Japan, 2020 edition (eighth revision); page last updated 28 August 2026. The leaf per 100 mL of water is calculated from the published infusion conditions and is not a figure written on that page. ⚠️ These four rows are not under identical conditions. Only hojicha and genmaicha share infusion conditions, so dropping the conditions column makes the table itself wrong.

The latte turned out to be a question of ratio

The latte turned out to be a question of ratio. The recipe Kurasu publishes in Kyoto dissolves hojicha powder in sugar and hot water to make a syrup, then combines it with warmed milk. Because the quantities are by weight, how much of what went into one cup can be read off the page. That the café cup and the home cup do not resemble each other is not a difference of skill. What to change to close the gap is chapter 2.

Roasting at home is more tangled

Roasting at home is more tangled. In the report of the Tea Research Station, Kochi Prefectural Agricultural Technology Center, the caffeine table of the following year (2014 data) shows almost no reduction at the light roast and finally 6–13% at the dark roast. No peer-reviewed paper on roasting in a household hōroku or frying pan has turned up within the range searched, either. Chapter 4 starts by settling conditions and measuring.