← Recipes

Traditional method

Mucor-type douchi

毛霉型豆豉 · 毛霉豆豉 · 豆豉 · Dòuchǐ · Mucor-fermented douchi · Mucor-type fermented soybeans · Mucor-type fermented black beans · Mucor-type dou-chi · Yongchuan douchi · 永川豆豉 · Tongchuan douchi · 潼川豆豉

Loading ratings…

A Chinese fermented soybean condiment from Chongqing and Sichuan, made by growing Mucor-type mould on cooked beans, then salting and maturing them into dark, savoury, soft beans with a concentrated aroma.

A small ceramic dish holding wrinkled dark brown-black finished Mucor-type douchi beans on a wooden surface.

What it is

Mucor-type douchi is a Chinese fermented soybean condiment made through a mould-led fermentation followed by salting and maturation. The Chinese name is 毛霉型豆豉 (máoméi xíng dòuchǐ), literally Mucor-type fermented beans. The finished product normally consists of individual dark soybean grains with a concentrated salty, savoury flavour. The beans may be soft enough to crush between the fingers but generally retain their shape. It is a cooking ingredient rather than a fermented bean dish usually eaten by itself.

Douchi is a larger family of Chinese fermented soybean products, not a single standardised fermentation. Scientific literature distinguishes Mucor-type, Aspergillus-type, Rhizopus-type and bacterial-type production. These categories describe the organisms or processes associated with the primary fermentation. They do not mean that every finished product contains only one organism. Mucor-type production is particularly associated with southwestern China, including Yongchuan in Chongqing and Tongchuan in Sichuan. Douchi from other places cannot be assumed to use the same mould, temperature or salting process.

The basic raw material is soybean, Glycine max. Black soybeans are especially associated with dark douchi, but Mucor-type methods also use yellow soybeans. During preparation, cooked beans are colonised by mould and become a fermented intermediate called douqu (豆曲). The mould produces enzymes that begin breaking down the soybean proteins and other substances. After this initial stage, salt and sometimes Chinese liquor, fermented rice or seasonings are added. The beans then mature, during which enzyme activity, other microorganisms and chemical reactions change the flavour further.

Mucor racemosus has been used in controlled experiments, including a study using strain CGMCC8700. Another study used Actinomucor elegans, a related mould in the order Mucorales, isolated from douchi koji. A broader survey of fermented soybean foods also identified Mucor circinelloides and other Mucorales. Finding a species in traditional beans does not make it an intentional or safe starter. The name Mucor-type describes a production tradition; it is not proof that a particular batch contains a pure culture.

The distinction matters when comparing recipes. An experimental M. racemosus process completed its initial mould stage at 25°C, whereas another Actinomucor process used temperatures below 15°C. Traditional regional douchi may take roughly a year to mature, while laboratory processes can be much shorter. Those are different methods with different organisms and outcomes, not interchangeable versions of one universal recipe.

Sources for this section

History

Fermented soybeans have a long documented history in China, but evidence for douchi generally must be distinguished from evidence for the specific Mucor-type process. Among the earliest secure records are fermented black soybeans, historically called shi (豉), associated with Western Han tombs at Mawangdui, near present-day Changsha in Hunan. Tomb No. 1 was sealed around 165 BCE. Its contents included jars and written labels identifying fermented soybean products. This demonstrates that salted fermented soybeans were known by that period. It does not identify the organisms used to ferment them or prove that the beans were manufactured using the same procedure as modern Yongchuan douchi.

The historical literature also contains references to shi in the Shiji, compiled during the Western Han, and later Chinese agricultural and food texts. The sixth-century agricultural treatise Qimin Yaoshu contains descriptions of fermented soybean preparation. These records are useful for establishing continuity in the broader practice of fermenting and preserving soybeans. They cannot establish continuity of a particular Mucor strain, because the microorganisms were not identified by the methods available to modern microbiology.

The regional history is more specific. China's national intangible cultural heritage record attributes the beginning of Yongchuan douchi production to approximately 1644, near the end of the Ming dynasty. The same record describes the local conditions and the use of Mucor-led production. For Tongchuan douchi, the national heritage record cites the Santai county chronicle and describes the Qiu family's arrival from Jiangxi around 1670, followed by the development of a local Mucor-koji technique. These dates are retrospective regional historical accounts. They should not be described as contemporary laboratory evidence or independently proven invention dates.

The Tongchuan heritage account also reports that the product was presented as tribute in 1678 and describes the later development of the family's sauce-making business. These details belong to the local tradition and rest on later accounts, not on an excavated object. Both Yongchuan and Tongchuan douchi-making techniques were included in China's second batch of national intangible cultural heritage entries in 2008, providing secure modern documentation of their recognised regional significance.

Microbiologists have since separated the traditional fermentation community from deliberately chosen starters. Work published in 2016 compared natural maturation with an Actinomucor starter and selected salt-tolerant organisms for the later stage. A 2019 study followed the metabolites of a Mucor racemosus fermentation, a 2020 study compared natural and artificial Yongchuan douchi, a 2025 survey described the bacterial and fungal communities of seven Sichuan and Chongqing products, and a 2026 study followed flavour change through a long maturation of Tongchuan douchi. Each describes one process under its own conditions, so none of them shows that a laboratory procedure reproduces every historical method.

Sources for this section

How it works: moulds, enzymes and chemistry

Mucor-type douchi begins with solid-state fungal fermentation. The cooked beans contain water, proteins, lipids and carbohydrates, providing nutrients for mould growth. A suitable Mucorales fungus germinates from its spores and develops branching hyphae over the soybean surface. The mycelium releases extracellular enzymes that begin digesting the substrate outside the fungal cells. The soluble products can then be absorbed by the mould or remain in the beans, where they contribute to subsequent flavour development.

Proteolysis is one of the central transformations. Soybeans contain substantial quantities of storage proteins, particularly glycinin and beta-conglycinin. Proteases cut peptide bonds within these proteins, producing progressively smaller peptides. Peptidases can break these fragments down further into free amino acids. Glutamic acid and aspartic acid are important contributors to savoury taste, while other amino acids become precursors for aroma compounds. The amount released depends on the strain, the available enzymes, substrate condition and duration of fermentation. Mould growth alone does not demonstrate that a particular enzyme has reached a useful concentration.

Other hydrolytic enzymes also affect the beans. Carbohydrate-degrading enzymes can release smaller sugars from accessible soybean carbohydrates. Phytase can break down phytic acid and release phosphate and inositol-containing products. In a 2019 Mucor racemosus experiment, myo-inositol increased substantially during fermentation, consistent with changes involving phosphorus-containing compounds. Several amino acids and putrescine also rose, while malic acid fell. These are figures from one fermentation and do not hold for all Mucor-type douchi.

After the mould-growing stage, salt changes the environment by lowering water activity and limiting the growth of many microorganisms. Salt does not immediately inactivate every enzyme, so chemical changes can continue during maturation. Liquor and fermented rice, when included, contribute ethanol and additional fermentation-derived compounds. Alcohols can interact with organic acids to produce esters, while amino-acid metabolism can generate aldehydes, alcohols and acids. Darkening and changes in aroma can also involve oxidation and non-enzymatic browning reactions, including Maillard chemistry.

Maturation is a mixed biological and chemical process even when the initial stage is mould-led. Surveys of finished douchi have detected Bacillus, Tetragenococcus, Weissella, yeasts and several fungal genera. A 2025 study of seven douchi products identified sixteen amino acids and one hundred volatile compounds. It found associations between particular microorganisms and compounds such as phenylacetaldehyde, 1-octen-3-ol and esters. Such correlations do not prove that the detected species produced those chemicals directly. Controlled inoculation and metabolite measurements are needed to establish individual microbial functions.

The changing aroma is therefore not a simple measure of fungal growth. Proteins first become soluble peptides and amino acids; later reactions consume some of those compounds while generating new flavour molecules. A 2019 experimental fermentation showed substantial metabolite changes during its early stages and comparatively little separation between samples after day fifteen. That finding describes the metabolic behaviour of one controlled M. racemosus process. It does not mean traditional douchi finishes maturing after fifteen days.

Sources for this section

How it is made and why each step matters

The first operation is selecting, cleaning and soaking the soybeans. Removing damaged beans and foreign material improves raw-material consistency. Soaking allows water to penetrate the beans so they can cook evenly. Cooking softens the bean structure, denatures proteins and reduces naturally occurring antinutritional factors. It also reduces the initial microbial population, giving the intended starter a better opportunity to establish itself. The cooking stage must be appropriate to the actual beans and equipment; soybeans that remain hard inside are not equivalent to thoroughly cooked beans.

The next stage is cooling and inoculation. The beans must be cool enough that the fungal inoculum is not damaged. In a 2019 experiment, soaked soybeans were steamed at 115°C for twenty minutes, cooled to 35°C and inoculated with Mucor racemosus CGMCC8700. The inoculated beans were incubated at 25°C for forty-eight hours. This produced the initial mould-fermented intermediate, after which 1% distilled wine and 8% sodium chloride were added and maturation continued at 25°C. Their distilled wine contained 45% alcohol before mixing. No household batch weight or reproducible inoculum quantity is given for it.

The two stages have different purposes. During initial incubation, fungal growth and enzyme production prepare the beans for further breakdown. After the addition of salt and liquor, conditions change: many microorganisms are inhibited, while enzymes, remaining microbial populations and chemical reactions continue altering the substrate. Maturation was followed for twenty-five days, and fifteen days was roughly the point beyond which the metabolomic profiles showed little further separation. This is a finding for that experimental system, not a safety-tested universal maturity endpoint.

Another published method used Actinomucor elegans rather than M. racemosus. In that study, soybeans were steamed at 100°C for 120 minutes, cooled below 40°C and inoculated with a prepared A. elegans culture. Initial cultivation continued for fifteen days below 15°C. The resulting douqu was combined with 12% salt, 5% fermented rice and 5% liquor, then matured in pottery jars for periods of one to five years. These conditions differ markedly from the faster M. racemosus method. They remain separate protocols.

Salt distributes through the beans during mixing and maturation, while liquor and fermented rice change the flavour environment. Even distribution is important because unevenly salted areas may support unwanted growth. Traditional regional methods can include further handling such as breaking up the mould-grown mass, managing surface growth, adding spices or seasoning, and storing the beans in sealed vessels. Such operations are specific to individual production traditions, not universal instructions for every Mucor-type douchi.

Neither laboratory method is a home recipe. The starter preparation, inoculation rate, moisture and water-activity limits and acceptance criteria for the finished beans all have to be set for a given process, and copying the temperatures and times from a laboratory protocol does not make a spontaneous fermentation safe.

Sources for this section

Regional and historical variations

Yongchuan douchi is one of the best documented traditional Mucor-type products. It originates in Yongchuan, now a district of Chongqing Municipality. China's national intangible cultural heritage record describes a production sequence involving selection, cleaning, soaking, cooling, mould cultivation, mixing, filling vessels and further fermentation. The process traditionally includes prolonged maturation under changing temperatures. The heritage record describes fermentation lasting around a year, considerably longer than the short processes used in some controlled laboratory studies.

Traditional Yongchuan manufacture is associated with cooler seasonal conditions. Studies of natural-type Yongchuan douchi describe the initial mould cultivation as taking place in the colder months, approximately November through February. The commonly cited production organism is a Mucor mould, although taxonomic descriptions and natural microbial populations are not always identical between studies. An investigation published in 2020 found that natural-type Yongchuan douchi contained greater microbial diversity than the artificial-type product used for comparison. It also showed higher amino nitrogen, more umami-related amino acids and a broader set of detected volatile compounds.

Tongchuan douchi comes from Santai County in Sichuan, historically known as Tongchuan. It should not be confused with the modern city of Tongchuan in Shaanxi. The regional heritage account connects its manufacture to the Qiu family and describes mould cultivation followed by further processing and long maturation. Black beans are steamed and left to grow Mucor for about twenty days, then mixed with salt and rice wine lees and matured in large jars for about a year, with the jars turned several times a year; aromatic seasonings and dehydration also feature in the local account. Its use of Mucor koji is explicitly part of the recognised local technique.

Long maturation remains important to Tongchuan's flavour. A study published in 2026 compared samples fermented for six, twelve, eighteen and thirty-six months. Both volatile and non-volatile compounds changed from stage to stage. Longer maturation was associated with increased development of certain esters, pyrazines and other aroma components. The study demonstrates that a product's chemical profile can continue changing long after the initial mould fermentation, even when visible fungal growth is no longer the main activity.

Modern production has also introduced selected cultures and accelerated methods. Some artificial-type Yongchuan production that has been studied used Aspergillus-led cultivation rather than the traditional Mucor-led route. It produced a different microbial community and flavour profile, with a shorter maturation period. It should therefore be classified by its actual fermentation process, not assigned to the Mucor category simply because it uses the Yongchuan name.

Regional methods, industrial adaptations and laboratory protocols must be kept distinct. Mucor racemosus can be used in a controlled process at a temperature different from that associated with traditional winter production. Actinomucor elegans has also been used experimentally. The wider existence of Rhizopus-type or bacterial-type douchi shows that similar-looking finished soybeans can result from different microorganisms and production techniques. Their methods and historical claims should not be transferred automatically between categories.

Sources for this section

How it is eaten and used

Mucor-type douchi is primarily a seasoning. The finished beans contribute salt, savoury amino acids, fermentation-derived aromas and a concentrated soybean taste to cooked food. Unlike fresh soybean ferments served in comparatively large portions, mature douchi is generally used in smaller quantities alongside other ingredients. It can be added whole, crushed or chopped, depending on the dish and the texture of the finished product.

Yongchuan douchi is closely associated with Sichuan and Chongqing cooking. Its recorded uses include twice-cooked pork, fermented-black-bean fish preparations and pork ribs. Beans can be chopped with garlic, warmed in cooking oil or included in a sauce before the main ingredients are added. Heating distributes the flavour into the oil and surrounding liquid. The fermented beans provide their own salt, so the amount of additional seasoning in a dish depends on the product being used.

Fish and meat dishes commonly use douchi as a concentrated seasoning rather than as a separate garnish. In steamed preparations, crushed beans can be distributed over the main ingredient together with aromatics. In stir-fries, they can be mixed into a sauce or briefly cooked with garlic and other seasonings. The same principle applies to vegetables and tofu: douchi supplies fermented soybean flavour without needing a large quantity of whole beans. These are applications of the finished condiment, not additional stages in making the douchi itself.

The maturity and form of the beans affect their use. A comparatively dry product can be chopped into distinct pieces, while a softer or moister preparation may mix more readily into sauces. Some commercial products are plain fermented beans; others include chilli, oil or additional seasoning. Those prepared sauces are not equivalent to an unseasoned batch of fermented douchi, and their ingredient composition should be recorded separately.

Regional foods also differ in their balance of salt, aroma and sweetness. Natural and artificial Yongchuan douchi differ in amino nitrogen, free amino acids and volatile compounds, while aged Tongchuan douchi develops further flavour changes with longer maturation. Consequently, two packages labelled douchi may give noticeably different results in the same dish. A recipe developed around one regional product need not taste identical when prepared with another.

Mucor-type douchi is the prepared condiment. Black-bean fish, twice-cooked pork and stir-fries are dishes that use it, and chilli douchi or black bean garlic sauces are separate products whose added ingredients and processing differ from plain matured beans.

Sources for this section

Reading results, identifying problems and understanding safety

A successful Mucor-type douchi fermentation depends on the intended organism growing under suitable conditions, producing the necessary enzymes and being followed by controlled salting and maturation. During the initial stage, visible fungal growth may appear as a pale, filamentous covering on the cooked beans. Its appearance alone is not a reliable way to identify the species. Many unrelated moulds produce pale or hairy colonies, and the appearance of a culture can change as it develops. A wild mould-covered bean cannot be identified as food-grade Mucor racemosus simply by inspecting or smelling it.

The final beans are commonly dark, salty, fragrant and relatively soft while retaining a recognisable grain structure. Traditional descriptions of Yongchuan and Tongchuan douchi refer to dark colour, characteristic aroma and a texture that breaks down readily during eating. These are useful descriptions of established products, not laboratory safety criteria. Unexpected colours, abnormal surface growth, putrid odours, persistent unexplained slime or evidence of mishandling are reasons to reject an uncontrolled batch, but the absence of those signs does not establish that the food is safe.

One major problem is microbial contamination. A 2025 investigation examined fifty-six traditional fermented soybean-food samples from China, including douchi, furu and dajiang. It identified thirteen Mucorales species, with Mucor circinelloides and Mucor racemosus among the most frequently isolated across the combined sample set. Some members of the order are opportunistic human pathogens under particular conditions. Their detection in fermented food does not mean every isolate causes disease, but it demonstrates why identifying the organism and its strain matters.

Biogenic amines are another concern. Microorganisms can produce these compounds by decarboxylating amino acids released during protein breakdown. Histidine can become histamine, tyrosine can become tyramine, and other pathways can generate putrescine, cadaverine and phenethylamine. A 2018 study measured six common biogenic amines in fifteen douchi samples. Although total amine concentrations were below the study's cited overall threshold, some samples exceeded reference levels cited for individual compounds, including histamine and beta-phenethylamine. These results concerned douchi samples generally and should not be represented as prevalence figures specifically for Mucor-type production.

The Mucor racemosus metabolomics experiment also found substantial increases in putrescine and phenylethylamine. This illustrates an important trade-off: protein hydrolysis supplies amino acids that can improve taste, but some organisms can convert amino-acid precursors into undesirable amines. Starter selection, hygiene and control of the microbial community therefore affect safety as well as flavour. Laboratory testing is needed to determine the actual amine content of a batch. Smell, dark colour and a strong savoury taste cannot measure histamine or other biogenic amines.

Production faults can begin before salting. Uneven cooking, unsuitable moisture, excessive accumulation of heat or poor inoculation can allow unwanted organisms to compete with the intended mould. Later, uneven mixing may leave areas with insufficient salt. A traditional Yongchuan producer interviewed in 2025 specifically identified undercooked beans, uneven cooling and poorly distributed salt as practical causes of defective fermentation. A salt percentage or fermentation time taken from a laboratory protocol is not a universal safety limit without measurements of the final product.

A metabolite profile that stops changing after fifteen days, as it did in the 2019 Mucor racemosus fermentation, shows that the chemistry had largely settled under those conditions. It does not show that pathogens are eliminated, that toxins are absent or that the beans will keep. Sequencing likewise shows which organisms are present, not which of them are active or responsible for a given aroma compound. Soybeans remain an allergen after fermentation. Beans meant for eating or sale need an identified food-grade starter, a validated process, testing, hygienic handling and verified storage conditions. A batch with unidentified mould or other signs of failure is not tasted or rescued by adding salt.

Sources for this section

History

Salted fermented soybeans (shi) were buried in a Western Han tomb at Mawangdui sealed around 165 BCE, but that find does not identify the mould or the method. China's heritage records date Yongchuan douchi to about 1644 and the Qiu family's Tongchuan douchi to about 1670; both rest on later regional accounts, not on surviving seventeenth-century laboratory evidence.

Ingredients

Scaling: source batch only

Dried soybeans (Glycine max)Cleaned and sorted; the dry weight is the basis for the other percentages.
Amount not stated
No starting batch weight is fixed; the published method gives only percentages.
WaterTap water for soaking and for rinsing the mould-grown beans.
Amount not stated
Soaking water is three times the volume of the dry soybeans; the rinse volume is not fixed.
Actinomucor elegans inoculumPure-culture spore suspension of Actinomucor elegans.
Amount not stated
Used at 1% v/w. The strain designation and spore concentration are not given, so the inoculum cannot be reproduced from this method alone.
Salt (sodium chloride)Mixed with the rinsed mould-grown beans.
Amount not stated
10% of the original dry soybean weight. This is a recipe proportion, not a validated food-safety limit.
Fermented glutinous riceAdded together with the salt and spirit.
Amount not stated
5% of the original dry soybean weight. Its own preparation is not described.
Distilled spiritAdded together with the salt and fermented rice.
Amount not stated
5% of the original dry soybean weight. The type and strength of the spirit are not given.

Method

Risks and precautions · Standard risk
Unidentified or unsuitable fungal cultures

Mucor-type fermentation relies on particular fungi. Unidentified environmental moulds may be contaminants or unsuitable strains. Appearance alone cannot establish organism identity or food suitability.

Hazard source
Microbial contamination during spontaneous fermentation

Research on natural Yongchuan douchi detected diverse fungal and bacterial populations, including potentially undesirable microorganisms. Warm, moist cooked soybeans can support competing organisms during fermentation.

Hazard source
Biogenic amine accumulation

Douchi can contain histamine, tyramine, putrescine, cadaverine and other biogenic amines. A study of fifteen commercial douchi samples found elevated individual amine concentrations in some products. These observations concern douchi generally, not specifically the 2015 Actinomucor protocol.

Hazard source
Insufficient preservation or uneven salting

Salt inhibits microorganisms partly by reducing available water, but the final water activity depends on the complete formulation and processing conditions. A nominal salt addition does not independently establish product stability.

Hazard source
Soya allergen

Soybeans are a recognised food allergen. Fermentation does not establish that the finished food is suitable for people with soya allergy.

Hazard source

Use a correctly identified, food-appropriate Actinomucor elegans starter prepared under controlled conditions, rather than collecting unidentified mould from the environment.

Control source

Follow a validated soybean preparation and cooking process, with appropriate hygiene and protection against contamination before inoculation.

Control source

Control incubation conditions and verify that the intended starter has established itself without unacceptable contamination.

Control source

Distribute salt uniformly and establish final preservation conditions using a validated process and measurements, including water activity where appropriate.

Control source

Evaluate biogenic amine risks through appropriate starter selection, process control and analytical testing.

Control source

Establish microbiological specifications, storage requirements and shelf life before supplying the product for consumption.

Control source

This is a published laboratory production method and not a validated home-fermentation recipe.

Do not substitute wild mould, visibly mouldy food or an unidentified fungal culture for the specified starter.

Do not combine the temperatures and times of this Actinomucor elegans protocol with those of the separate Mucor racemosus CGMCC8700 experiment.

The salt percentage, temperature and maturation time are processing conditions, not established safety-critical limits.

No validated final water activity, microbiological acceptance criteria or shelf life exists for this method.

Appearance, aroma and flavour cannot show the absence of pathogens, toxins or biogenic amines.

A suspected contaminated batch is not tasted or rescued by adding more salt.

Contains soybeans, a recognised allergen.

  1. Select and wash the dried soybeans, removing foreign material and damaged beans.
  2. Soak the soybeans in three times their volume of tap water for 4 hours at 25 °C, then drain.
  3. Steam the drained soybeans at 100 °C for 2 hours.
  4. Cool the steamed soybeans until they are cool enough not to harm the inoculum; this protocol fixes no cooling temperature.
  5. Inoculate the cooled soybeans with a pure-culture spore suspension of Actinomucor elegans at 1% v/w. Use an identified laboratory culture, never mould collected from food or the environment.
  6. Incubate at 28 °C and 90% relative humidity for 10 days. The mould-covered beans are the douqu (豆曲).
  7. Wash the douqu with tap water.
  8. Mix the washed douqu with salt equal to 10% of the original dry soybean weight, fermented glutinous rice equal to 5% and distilled spirit equal to 5%, spreading them evenly through the beans.
  9. Transfer the mixture to bottles and seal them.
  10. Mature the sealed bottles at 28 °C for 150 days.
  11. Treat the finished beans as an experimental product: the protocol sets no microbiological release criteria, shelf life or storage instructions.

Conditions

Fermentation160 days
Temperature28°C
Temperature controlIncubated

Places

China

Traditional preparation · Country evidence

Yongchuan District

Traditional preparation · Locality evidence

Santai County, historical Tongchuan

Traditional preparation · Locality evidence

Related

Community

No community contributions have been published for this entry yet.

Share what you made

Sign in to contribute. Published comments and photos are free to read.

Sources