# Koji fermentation

> Growing a mould on cooked grain so that its enzymes ferment food

- Date: c. 8,000 BCE
- Era: Prehistory
- Domain: [Food & Preservation](https://lastcodex.org/en/domains/food/)
- Rebuild stage: II (Settlement)
- Difficulty: 3/5
- URL: https://lastcodex.org/en/inventions/koji-fermentation/

Koji is steamed rice, barley, wheat or soybeans on which a mould, most often Aspergillus oryzae, has been grown until it is full of enzymes. Mixed with more grain, beans, salt or water, those enzymes turn starch into sugars and proteins into smaller pieces, the first step of sake, miso and soy sauce. Moulds grown on rice were already used to ferment a drink at Shangshan, in southern China, some 10,000 to 9,000 years ago.

![Koji fermentation · Growing a mould on cooked grain so that its enzymes ferment food](https://lastcodex.org/_astro/plate.B1s48wPf_mV0lv.webp)

## Legend

1. Shallow wooden tray
2. Layer of steamed rice grains
3. White felt of mould threads over the grains
4. Threads growing into a grain (magnified)
5. Spore-bearing head of Aspergillus oryzae

## What it is

Koji is grain or beans that a mould has colonised. In Japan the word means both the mould and the cooked rice, barley, wheat or soybeans it grows on. The mould is most often _Aspergillus oryzae_. It is grown on purpose in thin layers kept warm and very damp, and harvested before it turns green with spores. What the maker wants from it is its enzymes, the proteins that a mould releases to digest its food.

Koji is the starter of many East Asian foods and drinks. Rice koji mixed with steamed rice, water and yeast becomes sake. Soybeans mashed with koji and salt and left to mature become miso. Soybeans and wheat grown over with koji, then soaked in brine, give soy sauce. In China, ferments of the same family are called _qu_. They are often pressed into bricks of crushed grain and carry several moulds, yeasts and bacteria at once; some rely on a red mould, _Monascus_, the mould of red koji.

Traditional koji is made in a warm room lined with wood. Steamed rice is cooled, sprinkled with a powder of dried spores and spread in shallow wooden trays stacked on shelves. Workers turn and loosen it by hand while it grows.

## How it works

Steaming prepares the grain. Raw starch is packed in hard granules that enzymes attack only slowly; heat and water swell and soften them, so the starch becomes easy to digest. The rice is then cooled before the spores are added, because heat would kill them.

The spores sprout fine threads, called hyphae, that run over the surface of each grain and push into it, and the grains are soon covered with a white felt. The growing mould breathes and gives off heat, so the layer warms up by itself; turning it by hand lets the heat out and brings in air. As the grains dry a little, the mould slows down and stops. By then the threads have released large amounts of amylases, which cut starch into sugars, and proteases, which cut proteins into peptides and amino acids.

Makers choose their strain for the job. A sake koji needs a mould that makes plenty of amylase, since the sugar it frees is what the yeast turns into alcohol. A soy sauce koji needs more protease than amylase, because the flavour comes from broken-down soy protein; some soy sauce is made with a close relative, _Aspergillus sojae_. Left to grow on, the mould raises upright stalks ending in round heads covered with chains of yellow-green spores. Those spores are dried and sold to start the next batch.

![Large wooden vat bound with plaited bamboo hoops on an earth floor, part of its wall cut away to show the reddish-brown miso paste inside, with a flat wooden lid on the paste weighed down by a pyramid of large rounded river stones](https://lastcodex.org/_astro/v1.j1xQcVV7_29S230.webp)
_Miso matures under a lid weighed down with stones._

The work of the koji goes on after it has been mixed in. In miso, its enzymes keep digesting the soybeans while the paste matures, and microbes that tolerate salt join in: a yeast, _Zygosaccharomyces rouxii_, and a lactic acid bacterium, _Tetragenococcus halophilus_. The salt holds back most other microbes, and the paste darkens and gains flavour with time. Rice miso can be ready fairly soon, while soybean miso matures for months, and longer still without heat treatment.

_Aspergillus oryzae_ is a domesticated mould. Its genes are almost identical to those of _Aspergillus flavus_, a common mould of crops that can make aflatoxin, a strong poison. Japanese studies of starter strains in the 1970s found that none of them produced aflatoxin, and genome work has since shown that the genes needed for it are partly deleted or silent in koji strains.

## How to do it

Growing koji is not a home project: a warm, damp tray of cooked rice welcomes wild moulds just as well, some of them poisonous, and a beginner cannot tell them apart. Dry rice koji, sold in some food shops, shows its main enzyme at work without growing anything.

![Two white saucers side by side on a wooden board, each holding a small pool of translucent starch paste; in the left one a drop of iodine has turned deep blue-black, in the right one, mixed with a few whitish grains of rice koji, the iodine stays light brown; a glass dropper lies in front of them and a small stoppered brown bottle stands behind](https://lastcodex.org/_astro/v2.BX2-W1qm_DSLzp.webp)
_Iodine turns blue-black where starch remains and stays brown where koji has broken it down._

Stir a spoonful of cornflour or potato starch into a little cold water. Ask an adult to pour hot water from the kettle over it while you stir, until it turns into a thin, clear paste, and let it cool until it is barely warm. Put a spoonful of paste in each of two saucers. In the second, mix in a pinch of koji grains crushed with a few drops of lukewarm water. Cover both saucers and wait.

Then add a drop of iodine solution to each. Iodine is yellow-brown on its own and turns blue-black when it meets starch. The plain saucer turns dark at once. In the saucer with koji, where the amylase has been cutting the starch into sugars, the colour stays lighter, and after a longer wait it stays brown. To check that an enzyme is at work, an adult can first boil a pinch of koji in a little water: heat destroys the enzyme, and that saucer turns blue-black like the plain one.

## History and evolution

The oldest traces of moulds grown on grain to ferment it come from the lower Yangtze region of China. At Shangshan, in Zhejiang, pots from the earliest phase of the site, some 10,000 to 9,000 years old, hold residues of rice together with remains of yeasts and of the red mould _Monascus_. The researchers read them as traces of a _qu_ starter, moulds grown on rice to brew a drink. At Qiaotou, a site about 9,000 years old in the same province, pots for drinking held moulds whose shape matches _Aspergillus_ and _Rhizopus_, identified by their form rather than by species. Machida and his colleagues, writing on the genetics and history of koji, place the origin of this kind of fermentation in China 3,000 to 2,000 years ago, much later than these residues.

![A brick of pressed crushed wheat with one corner broken off and lying beside it, its surface and the broken face speckled with white and yellow-green mould, on a straw mat with two whole bricks behind it](https://lastcodex.org/_astro/v3.CJ0hj3m4_Z21AVuT.webp)
_Chinese ferments are often pressed into bricks of crushed grain in which several moulds grow._

Chinese scholars of the 2nd century CE already wrote of pastes of soybeans, meat or fish made with grain ferments. A farming manual of the 6th century, the _Qimin yaoshu_, gives a soybean paste made with ferments of powdered and whole wheat, salt and brine. These pastes, _jiang_ in Chinese, became _hishio_ in Japan, and in the 8th century a palace office attached to the imperial kitchen made hishio from soybeans.

In Japan the starter became a trade. In the Muromachi period, from the 14th to the 16th century, sellers of koji spores were organised in a guild, the Koji-za. Spore makers grew their cultures with the white ash of burnt hardwood leaves, which held back unwanted microbes. In the 1870s Hermann Ahlburg, a German teaching in Tokyo, isolated the mould from koji and named it _Eurotium oryzae_; Ferdinand Cohn later renamed it _Aspergillus oryzae_.

The mould then entered industry. In 1894 Jokichi Takamine patented a way of growing it in thin layers on damp wheat bran and extracting its starch-digesting enzyme, sold as Taka-diastase, a remedy for poor digestion. Its genome was sequenced in 2005. Comparisons with _Aspergillus flavus_ point to a mould domesticated by long selection in the hands of starter makers, which has lost the ability to make aflatoxin and makes enzymes in great quantity.

## Key facts

- Oldest mould ferment found: Shangshan, China, about 10,000 to 9,000 years ago
- Main koji mould: Aspergillus oryzae, a domesticated relative of Aspergillus flavus
- First isolated from koji: By Hermann Ahlburg, in Tokyo, in the 1870s
- First enzyme product from the mould: Taka-diastase, patented by Jokichi Takamine in 1894

## Requires

- [Pottery](https://lastcodex.org/en/inventions/pottery/): Clay turned to stone by fire (20,000 years ago)
- [Harvest and threshing](https://lastcodex.org/en/inventions/harvest/): From standing wheat to stored grain (15,000 years ago)

## Sources

- [Genomics of Aspergillus oryzae: learning from the history of Koji mold and exploration of its future (DNA Research 15, 2008)](https://pmc.ncbi.nlm.nih.gov/articles/PMC2575883/), Oxford University Press
- [Japanese Traditional Miso and Koji Making (Journal of Fungi 7, 2021)](https://pmc.ncbi.nlm.nih.gov/articles/PMC8307815/), MDPI
- [Evolution of Aspergillus oryzae before and after domestication inferred by large-scale comparative genomic analysis (DNA Research 26, 2019)](https://pmc.ncbi.nlm.nih.gov/articles/PMC6993814/), Oxford University Press
- [Identification of 10,000-year-old rice beer at Shangshan in the Lower Yangzi River valley of China (PNAS 121, 2024)](https://pmc.ncbi.nlm.nih.gov/articles/PMC11665889/), National Academy of Sciences
- [From Artisan Experience to Scientific Elucidation: Chinese Traditional Fermentation Starters (Qu) (Foods 14, 2025)](https://doi.org/10.3390/foods14223814), MDPI
- [Early evidence for beer drinking in a 9000-year-old platform mound in southern China (PLOS ONE 16, 2021)](https://pmc.ncbi.nlm.nih.gov/articles/PMC8360526/), PLOS
- [Remains of 9,000-Year-Old Beer Found in China](https://www.smithsonianmag.com/smart-news/remains-9000-year-old-beer-found-china-180978563/), Smithsonian Magazine
- [The History of Shoyu (Soy Sauce) 1, by Ryoichi Iino (Food Culture 1)](https://www.kikkoman.com/en/kiifc/journal/pdf_01/e_012_015.pdf), Kikkoman Institute for International Food Culture
- [Process of making diastatic enzyme (US patent 525,823, J. Takamine, 1894)](https://patents.google.com/patent/US525823A/en), Google Patents
- [Detecting starch in food on a microscale](https://edu.rsc.org/experiments/detecting-starch-in-food-on-a-microscale/563.article), Royal Society of Chemistry
- [Food Tests (WJEC GCSE Biology): Revision Note](https://www.savemyexams.com/gcse/biology/wjec/18/revision-notes/1-cells-organ-systems-and-ecosystems/1-3-digestion-and-the-digestive-system-in-humans/food-tests), Save My Exams
- [Hatcho miso: interview with Maruya Hatcho Miso](https://japan-food.jetro.go.jp/odan/sg/miso/index.html), JETRO, Japan External Trade Organization
- [The history of soy sauce](https://www.soysauce.or.jp/en/the-history-of-soy-sauce), Japan Soy Sauce Brewers' Association
- [History of Miso and Soybean Chiang, page 2](https://www.soyinfocenter.com/HSS/miso2.php), Soyinfo Center
- [Koji Starter and Koji World in Japan (Journal of Fungi 7, 2021)](https://pmc.ncbi.nlm.nih.gov/articles/PMC8304044/), MDPI
- [GCSE Biology teacher guidance: investigating the effect of amylase on starch with iodine](https://www.eduqas.co.uk/media/z5gbqqcq/wjec-gcse-biology-tg-07-11-2019.pdf), WJEC/Eduqas

Fact-checked on 2026-10-21.

---

The Last Codex · https://lastcodex.org/en/inventions/koji-fermentation/ · © The Last Codex. Texts, plates and figures: CC BY-NC-ND 4.0.