a fungus that grows out of a caterpillar

01Two fungi and one name

Today the word cordyceps means several things at once: a genus of fungi, an expensive Himalayan natural product and capsules at the pharmacy. Before the history begins, it is worth untangling these things; otherwise the story gets knotted before it has even started.

What caterpillar fungi have in common is that they are parasites of insects. A spore catches on a caterpillar or a pupa, germinates, penetrates the body and fills it with mycelium. The host dies but does not fall apart like other carcasses: it stays permeated by the fungus, firm, almost as if turned on a lathe. In time a fruiting body grows out of it, a small stalked club in the upper part of which spores ripen in tiny sacs. The whole scene, half animal and half plant, looks like something invented by a poet with a weakness for parables. In reality it is ordinary fungal work, done in nature by a great many related species, each with its own hosts.

Two of them play the main roles in this story. They are not the only caterpillar fungi in the world, just the ones around which the most history and the most money revolve.

The Chinese caterpillar fungus, today Ophiocordyceps sinensis, is a highlander. It grows on the Tibetan Plateau and in the surrounding mountains, according to the IUCN assessment only above roughly 3,500 metres¹. It infects the caterpillars of moths of the genus Thitarodes (formerly Hepialus), which live in the soil of mountain meadows². Tibetans call it yartsa gunbu, “summer grass, winter worm”; the Chinese name 冬虫夏草, “winter worm, summer grass”, reverses the word order³. Both names describe what can be seen in the meadow: in summer a dark stalk sticks up out of the grass, and when it is carefully dug out, a caterpillar hangs from it.

The scarlet caterpillarclub, Cordyceps militaris, is at home across the whole northern hemisphere, in Europe, Asia and North America⁴. It infects the pupae and caterpillars of a wide range of moths and butterflies⁵, and its bright orange clubs grow out of pupae buried in the soil, in Europe from August to November⁴. It is the type species of the genus Cordyceps: anyone who wants to know what a “true” cordyceps looks like has to look at this one⁶.

A third fungus plays the part of a stowaway in this story. The fermentation strain Cs-4, from which cheap mycelium for capsules is made, is, according to taxonomists, Paecilomyces hepiali, since 2020 Samsoniella hepiali. It is a different species, and it belongs to the family of the scarlet caterpillarclub, not to that of the Chinese caterpillar fungus⁷. On labels it nevertheless appears as “Cordyceps sinensis Cs-4” or, in Czech, as “extrakt z mycelia housenice čínské”, an extract of Chinese caterpillar fungus mycelium⁸. In 2025 a representative of a European company summed it up for the trade press quite candidly: scientifically it is inaccurate, but commercially it has caught on⁹.

Why does it matter? Because what holds for one fungus need not hold for another. A different species has different genes, different substances and a different history. It is like surnames: two people called Smith need not be related, let alone have the same illnesses. The tradition of the Himalayan fungus cannot be stitched onto a fungus grown on rice, and the result of a study with fermented mycelium is not a result for fruiting bodies. The whole story that follows is really an exercise in telling things apart.

↑ back to contents

02Summer grass, winter worm

If the Chinese caterpillar fungus had a birth certificate, the name on it would be Tibetan. The oldest text that researchers agree on is attributed to the Tibetan physician and lama Zurkhar Nyamnyi Dorje (Zur mkhar Mnyam nyid rdo rje), founder of the Zur medical tradition. He lived from 1439 to 1475³. His short treatise on yartsa gunbu, titled An Ocean of Aphrodisiacal Qualities in its English translation, survives in a collection of medical instructions whose modern edition was published in Lanzhou in 1993^{18, 19}. The exact year of writing is not known. If the attribution is correct, the text was written during the author’s lifetime. The researcher Daniel Winkler calls it the oldest known record of this fungus, and the most detailed Chinese review of the sources, from 2024, assesses it in the same way^{3, 20}. The oldest record found, however, does not say when the fungus began to be used; it says only which source documents it.

What does the treatise say? According to Jakob Winkler’s English translation, the fungus tastes sweet and a little astringent, is oily and slightly warm, dispels “prana” and bile diseases and does not increase phlegm. Above all, it “increases especially semen”: it serves sexual desire, offspring and vitality alike¹⁹.

The recipe is a little poem of practical pharmacy. The root is to be picked at the right time, ground to a fine powder and mixed with yeast. The chest of a sparrow is filled with this mixture, and the sparrow is boiled. Then the chest is opened with a golden instrument, a litre of sheep’s milk is added, and it is all boiled in an iron pot until the liquid is gone. Black pepper, bezoar, bear bile and other ingredients also go into the mixture, and, when obtainable, a “snow lizard”; sugar as well. From this, pills the size of sheep droppings are rolled. They are taken at dawn with arak or strong beer for a whole month, and during that time the patient is to abstain from lovemaking, raw food and spoiled vegetables¹⁹. This is a historical recipe, not a verified treatment. But anyone who expected a quiet little herb for coughs is caught off guard: the oldest documented use of the Chinese caterpillar fungus is as a tonic and an aphrodisiac.

Older Tibetan mentions are disputed. The Four Tantras, the foundational work of Tibetan medicine written between the 8th and 11th centuries, mention a medicine called daji, which some Tibetan scholars identify with yartsa gunbu. Winkler calls this identification controversial³. Chinese secondary literature goes further still and speaks of Tibetan translations from the 8th century. Chen Siqi’s review records them but points out that they contradict the generally accepted 15th century; an assessment by Tibetologists could not be found²⁰.

The great Chinese pharmacopoeias, the so-called materia medica, are silent about the Chinese caterpillar fungus for a long time. It is unknown to the Zhenglei bencao of 1082 and to Li Shizhen’s famous Bencao gangmu of 1590–1596; both works record a different insect fungus, the “cicada flower”²¹. Chen Siqi’s review regards a gazetteer of Sichuan province from the reign of the Yongzheng Emperor (1723–1735) as the first explicit Chinese record, and the gazetteer frankly admits: the materia medica do not record it. Then it adds that the fungus is warm in nature and replenishes essence and marrow²⁰. In 1744 the man of letters Yuan Dong describes a gift from a friend from Shaanxi: the fungus steeped in wine, which, when drunk, wards off illness and prolongs life²⁰.

The first detailed pharmacopoeia entry is held to be the one in the Ben cao cong xin, written in 1757 by the physician Wu Yiluo: the fungus is sweet and neutral, protects the lungs, benefits the kidneys, stops bleeding, dissolves phlegm and ends exhausting coughs^{20, 22}. Here, in the surviving sources, stands the “lung and kidney” profile that every seller knows today. Between the Tibetan treatise and the Chinese entry lie roughly three centuries and something like a change of career: a remedy for strength and fertility had become a cough medicine.

Wang Ang’s Ben cao bei yao of 1694 is often given as the first Chinese record³. Chen Siqi’s review, however, points out that the passage on the caterpillar fungus is a later addition to this work, and the Beijing medical products authority regards 1757 as the year of the first record^{20, 22}. Two small details also speak for this. The gazetteer of 1723–1735 states explicitly that the pharmacopoeias do not know the fungus, and the text attributed to 1694 matches the 1757 entry word for word. Only a look at the first edition can decide. Until then, 1694 carries a question mark.

In the second half of the 18th century the fungus is written about more and more. Zhao Xuemin, in his supplement to the famous pharmacopoeia, written in about 1765–1803, mentions a duck stuffed with the fungus that equals one liang of Korean ginseng in strength²¹. That duck is worth remembering: Paris knew about it a generation earlier.

The word order is a clue as well. The Tibetan name says “summer grass, winter worm”; the standard Chinese order reverses it. Several early Qing texts, however, still write “summer grass, winter worm” the Tibetan way, and so does a French transcription from 1726^{20, 23}. It looks as if the fungus moved from Tibet to China name and all, and the name was only rearranged into the Chinese order over time. This is only a conjecture from comparing the sources; the sources themselves do not say so.

Trade between Tibet and China is documented from at least the 17th century. The fungus was exchanged for tea and silk and also served as currency³. In the 19th century it reached Canton in little bundles tied with silk, about a dozen pieces each²⁴.

↑ back to contents

03Legends without contemporary evidence

Stories gather around every rare thing, and the Chinese caterpillar fungus is rare enough. Some of them are repeated in books and on packaging, so it is worth going through them one by one. They are stories, often beautiful ones, for which no contemporary evidence has been found so far.

Two thousand years. The claim that the fungus has been used for “two thousand years”, or even “since 2000 BC”, appears in popular texts and also in one Korean scientific review, which offers no evidence for it²⁸. Most often it rests on a lacquered case from the tomb of the Marquis of Haihun, from the time of the Western Han dynasty, which is said to have contained a “suspected caterpillar fungus”. That identification, however, has no citation; its only source is a remark made by one entomologist on a discussion forum in 2017²⁰. Just as uncertain is the “ice silkworm” from a 4th-century collection of curiosities: it measures seven inches, is black and has horns and scales, and not even the Chinese review that mentions it is sure that it is this fungus²⁰.

The Tang dynasty. An alleged mention from the year 620 has been repeated in the literature for decades. Winkler calls it unsubstantiated, and the chapter of a 2024 scientific book that gives it has no citation for it^{3, 2}.

The empress and the duck. The prettiest legend tells of the ageing Empress Wu, who in 690 was troubled by a cough that grew worse in winter and that not even the court physician’s expensive medicines helped. The court cook, Master Kong, remembered that back home the caterpillar fungus was stewed with chicken, and to be on the safe side he cooked it with a duck. The empress, however, saw dark worm-like things in the soup, concluded that the cook wanted to kill her and had him locked up. His friend, Master Li, wanted to prove Kong’s innocence and came up with a trick: he stuffed the fungus down the duck’s throat so that it could not be seen. The empress ate the soup twice a day and after a month stopped coughing. After that, all that remained was to explain what had been in that first soup². The story is a nice one, but there is no source for it: the book that tells it gives none, and the review of Chinese sources does not know it. Moreover, the oldest Chinese records of the caterpillar fungus that could be traced are roughly a thousand years younger. Most likely this is a projection back in time of the Qing recipe with the duck, which is documented²⁰.

The Qianlong Emperor. According to another story, in 1764 the Qianlong Emperor suffered from dizziness and severe pain in the lower back, and the court physicians were at a loss. Help came only from a folk healer with a drink based on the caterpillar fungus, and the emperor then supposedly invited thousands of people over eighty to a “feast of a thousand elders” so that they could taste it too. The book that tells this refers to a source that is missing from its own bibliography².

Tribute for the court. It is often written that the fungus had to be delivered to the Qing court as a compulsory tax. No evidence for this could be found. All that is documented is what the Jesuits wrote: that the fungus is so rare that it can be seen almost only in the palace^{29, 24}.

In the actual sources, the Chinese caterpillar fungus is a novelty in China at the turn of the 17th and 18th centuries. That takes nothing away from it; if anything, the opposite. Its documented journey from the Tibetan mountains through a provincial gazetteer into the pharmacopoeia, and from there to Paris, is more adventurous than any of the legends.

↑ back to contents

04Paris 1726: a caterpillar’s shed skin

The Chinese caterpillar fungus came to Europe by post. The Jesuit Dominique Parrenin, who lived in Peking and died there in 1741, sent the Paris Academy of Sciences (Académie royale des sciences) what its record calls a considerable gift, very much to its taste, together with two letters from August and October 1723. The Academy entered the gift in its volume for 1726³⁰.

That volume says that the Chinese regard the thing as a plant in summer and a worm in winter. The academicians examined it. They saw the perfect shape of a worm, dry and yellowish, nine lines long, that is, about two centimetres. And they decided that it was the true shed skin of some caterpillar. According to them, the caterpillar gnaws off the end of a rootlet, pushes its rear end into the hollow so that it can moult more easily, and the skin then stays there. The naturalist René-Antoine Ferchault de Réaumur, a great expert on insects, fully satisfied himself of this, according to the record³⁰.

Réaumur also wrote a paper of his own on the “plant-worm”; the margin of its first page bears the date 21 August 1726. Right at the start he recalls the gratitude the Academy owes “Father Parennin” for his splendid gifts, and then repeats what the Jesuit had written: the plant is foreign to China and very rare, it can be seen almost only in the palace in Peking, and it grows in Tibet and a little also on the borders of Sichuan province²³.

It is a fine example of how an expert sees: attentively, thoroughly and in the light of what they already know. Réaumur knew caterpillars, and so he saw a caterpillar. There is no reason to laugh at him: in 1726, recognising a fungus in something that looks like a dry caterpillar would have given anyone trouble.

Parrenin himself revealed more about the medicinal side of the matter. His report was printed by Jean-Baptiste Du Halde in his great description of China, published in Paris in 1735 and in The Hague in 1736²⁹. According to it, the fungus resembles ginseng in its effect but does not cause bleeding, and it restores strength lost through overwork or long illness. And the recipe is a culinary one: five drachms of the fungus are put inside a domestic duck, which is cooked slowly and eaten morning and evening for eight to ten days. Parrenin tried it on himself when a journey had exhausted him; both the fungus and the recipe had been given to him by the governor-general of Sichuan and Shaanxi. According to the report, a poor-quality sample was paid for with four times its weight in silver²⁹. So Paris knew the duck recipe before Zhao Xuemin wrote it down in China.

Incidentally, two errors keep being repeated in the literature. Parrenin did not bring the specimens to France: he died in Peking, and the Academy recorded a consignment from China³⁰. And Du Halde was not the one the fungus put back on his feet. He compiled and published the missionaries’ reports; the one treated was Parrenin²⁹.

The historian Lu Di traces how the fungus travelled on: to Britain around 1831, to Russia in 1851 and to America in 1891. According to him, it thereby also influenced European research on fungi that parasitise animals²⁷.

About a hundred years later, the fungus inside was already being written about. An English-language natural history of insects, whose 1837 edition is available in the Internet Archive, retells Parrenin’s report and the Paris explanation about moulting, and then adds that when the larva is cut open, the root of a fungus is found inside, filling its whole body from the head to the other end³¹. The fungus received its scientific name in 1843, when the British mycologist Miles Joseph Berkeley described it as a separate species, Sphaeria sinensis; his specimens were still immature^{11, 32}.

↑ back to contents

05A hundred years of names

Fungi are named according to where scientists happen to place them, and scientists change their minds. The scarlet caterpillarclub has been through this like few other fungi.

In the same years that the Paris Academy was examining the Chinese “plant-worm”, a catalogue of the plants around Paris by the botanist Sébastien Vaillant came out (1727), containing an illustration of a small club-shaped fungus. A quarter of a century later Carl Linnaeus relied on that illustration when he described the species Clavaria militaris in 1753^{17, 6}. He placed it among the club fungi, a completely different group; most likely he was guided by the club shape. So within a few years Paris saw both main characters of this story, without anyone suspecting that they were related.

Then mycologists moved it from genus to genus. It was a Sphaeria (1792), a Hypoxylon and a Xylaria (1821), a Corynesphaera (1822), a Kentrosporium (1844) and a Torrubia (1865)³². Meanwhile, in 1818, the Swedish mycologist Elias Fries established the genus Cordyceps for it. That genus, too, had a difficult birth: the original printing carried the name Cordylia, which Persoon had already used in 1807, and on the replacement leaves of the book, known technically as cancellans, it became Cordyceps¹⁶. In 1823 Fries himself demoted his genus to a mere section of Sphaeria, and it was not revived until Link did so in 1833¹⁶. Authors have their weak moments too. The valid name today is Cordyceps militaris (L.) Fr., and the scarlet caterpillarclub is the type species of its genus⁶.

The Chinese caterpillar fungus followed a similar path, only later. Berkeley described it in 1843 as Sphaeria sinensis, the Tulasne brothers moved it to the genus Torrubia in 1865, and in 1878 Saccardo transferred it to the genus Cordyceps as Cordyceps sinensis³². It spent almost a hundred and thirty years under that name, and that is also the name most labels know it by.

Then came molecular biology. In 2007 an international team led by Gi-Ho Sung compared stretches of DNA from 162 fungi of this group and showed that the old genus Cordyceps was not a natural group; the features by which the fungi had been sorted until then were misleading³³. The team split the genus accordingly. The scarlet caterpillarclub stayed in Cordyceps, because it is the type species of the genus, and the genus became the basis of the newly defined family Cordycipitaceae. The Chinese caterpillar fungus moved to Ophiocordyceps, a genus established by Thomas Petch in 1931, and to the new family Ophiocordycipitaceae^{33, 6}. Since 2007, then, “Cordyceps sinensis” has not been a Cordyceps. As if the most famous member of the family had turned out to be a neighbour all along.

↑ back to contents

06The scarlet caterpillarclub: a European birth certificate

The research for this text looked for evidence that the scarlet caterpillarclub was used as a medicine before the 20th century, whether in China, Korea, Japan or Europe. It found none.

The authors who discovered cordycepin in 1950 open their paper with a sentence saying that previous work on this fungus had been confined to its form and structure³⁴. The Japanese collector Yahagi knows it as a wild fungus and as something to be cultivated, but writes nothing about medicinal use²¹. A review of Tibetan, Mongolian and Qing sources apparently does not mention it at all²⁰. Claims about “traditional Chinese medicine” or about use “since 2000 BC” are repeated without evidence, and where they do have evidence, it concerns a different fungus^{28, 2}. One chapter of a 2024 academic book describes the actual course of events quite accurately: natural specimens of the Chinese caterpillar fungus were scarce, so substitutes multiplied, the scarlet caterpillarclub among them, and cultivation developed².

Missing evidence is not in itself proof that no tradition existed. But two thousand years of medicinal use of the scarlet caterpillarclub could not be documented in the sources checked. Its scientific story, by contrast, can be traced in the records, from the Paris drawing to the genome.

You are more likely to come across it in the wild by chance, usually while looking at your feet for some other reason. Roman Maňák describes it as a species found scattered in meadows, damp grassy places and steppe grassland, less often at forest edges and in warm oak-hornbeam woods; it is abundant in southern Moravia, especially in the White Carpathians, and fairly common on the sands of the Lower Morava Valley⁵. The fruiting bodies grow from dead pupae in the soil, in Europe from August to November⁴. At the time of the research (October 2026), the public GBIF database held only five records from Czechia, the most recent from Libochovice, dated 29 September 2026³⁵. That is more likely a sign that nobody digitises the finds than that the fungus is really so rare.

The 2024 Red List of fungi of the Czech Republic, an expert assessment of how threatened individual species are, does not include the scarlet caterpillarclub. It does, however, include four Czech relatives of the Chinese caterpillar fungus from the family Ophiocordycipitaceae, among them two known in Czech as “housenec střevlíkový” (ground-beetle caterpillar fungus) and “housenice velkovýtrusá” (large-spored caterpillar fungus)¹⁵. Legal protection of species is set by law, not by this list³⁶. Czech mycology has also given these fungi scholarly attention, including in the Šumava: in 2002 Holec and Suková analysed finds of the fungus then listed as Cordyceps longisegmentis from relict pine forests³⁷.

↑ back to contents

07Glasgow 1950: the pupa that did not rot

Cordycepin did not come out of a pharmacy but out of a lab that was looking for antibiotics. It was after the war, when, following penicillin, the search was on for more substances that would stop bacteria, and the chemists in Glasgow were struck by a small detail that a collector would not notice: the body of the host enclosed in the fungus does not decompose. A pupa that would otherwise go mouldy and fall apart stays whole inside the fungus. It occurred to them that something in the fungus was protecting it, perhaps a substance that kills other microbes³⁴.

The culture was supplied by Dr S. A. Hutchinson. He had established it as a single-spore isolate from a fruiting body that someone had collected in the autumn of 1948 in Tollymore Park, County Down, Northern Ireland³⁸. The culture grew in a simple nutrient solution of glucose and digested milk casein, and from it the chemists obtained a new substance. They named it cordycepin and announced it on 2 December 1950 in a letter in Nature. The letter is signed by Cunningham, Manson, Spring and Hutchinson, in that order, not in the order in which they are often cited today³⁴. A year later Kenneth Cunningham defended a thesis with the hopeful title Cordycepin, A New Antibiotic³⁸.

The chemistry of the day did not yet have today’s instruments, and the structure proposed in Glasgow was wrong: it assumed an oddly branched sugar, which was called cordycepose³⁸. The correct answer came only in 1964. Cordycepin is 3′-deoxyadenosine: adenosine, one of the basic building blocks of living cells, with an oxygen missing at one position³⁹. In the same year E. A. Kaczka’s team also found it in the common mould Aspergillus nidulans⁴⁰. So within a single year, an exotic fungal substance became a nucleoside that even a mould can make.

The missing oxygen is exactly what made cordycepin a useful tool. The cell mistakes it for adenosine and builds it into a growing RNA molecule; because the hook for the next link is missing at that position, the chain stops³⁹. Biologists made use of this. In 1970 they used it to tell apart different kinds of RNA in human cells grown in the lab⁴¹. Cordycepin became a reliable laboratory helper before anyone started writing about it in advertisements.

As a medicine, it was a harder case. In the human body it is rapidly broken down by the enzyme adenosine deaminase, and it gets into cells poorly; a 2021 oncology study therefore ascribes to it “minimal in vivo activity”⁴². Doctors tried giving it intravenously together with pentostatin, a substance that blocks that enzyme. In 1997–2000, 14 patients with acute lymphoblastic leukaemia received it in this way; according to a later annual report by OncoVista, the company that took up this research, no adverse effects associated with cordycepin occurred, but no peer-reviewed result was published⁴³. A follow-up study was stopped in 2009, after five patients, for lack of money^{43, 44}. Today NUC-7738 is being tested: a synthetic compound derived from cordycepin and modified to get around the enzyme. It is given intravenously to patients with advanced tumours; in August 2026 the company announced that recruitment for the second phase of the study was complete and promised final data by the end of the year^{42, 45}.

One simple thing follows from this. When a pure compound or a synthetic modification of it is given intravenously to cancer patients, that says nothing about what the fungus will do in a tea or a capsule. The substance, the dose, the route into the body and the people who receive it all differ.

And for a European lab, the nicest detail of the whole story is this: the culture from which cordycepin was first isolated came from a wild European fruiting body from a park in Northern Ireland.

↑ back to contents

08The questions change: from “what is it?” to “how does it work?”

The history of the science of caterpillar fungi can be read as a series of questions, each taking over from the last. In the 18th century science asked: is it a plant or an animal? In the 19th century: where does it belong and what is it called? In the mid-20th century: what does it make? And at the start of the 21st century: how does it work, and why does it work one time and not another?

Genomes answer that last question. In 2011 a Shanghai group led by Wang Chengshu read the genome of the scarlet caterpillarclub⁴⁶. It is small, about 32 million letters, and it contains no genes for known mycotoxins dangerous to humans⁴⁶. That is good news, even though on its own it does not prove that any particular product is safe. The genome also revealed a biological peculiarity. The scarlet caterpillarclub has two mating types, something like two sexes, and usually needs both; in exceptional cases, however, it can form a fruiting body even without a partner⁴⁶.

The genome of the Chinese caterpillar fungus is another world. It has about 116 million letters, inflated by repetitive stretches, and by one estimate this inflation took place 38 million years ago, at the same time as the uplift of the Tibetan Plateau⁴⁷. According to a genome study, its sexual reproduction is triggered only under natural environmental conditions⁴⁸. That offers one explanation of why it could not be cultivated for so long.

The big turning point came on 21 December 2017. Xia and colleagues showed that the scarlet caterpillarclub makes cordycepin and, alongside it, pentostatin, and that both compounds come from a single cluster of four genes⁴⁰. Pentostatin acts as cordycepin’s bodyguard: it blocks the enzyme that would break cordycepin down. Only when so much cordycepin builds up that it would harm the fungus itself does the guard step aside, and the excess is broken down⁴⁰. The strain of the Chinese caterpillar fungus in their experiments did not make cordycepin, and the genes of this cluster were not found in the genomes of the Chinese caterpillar fungus that have been read. The authors state explicitly that this contradicts earlier claims that the Chinese caterpillar fungus produces cordycepin⁴⁰. Chemical analyses of natural fruiting bodies of the Chinese caterpillar fungus found little or none of it⁴⁹. Cordycepin was, on the other hand, produced by yet another related fungus, Cordyceps kyusyuensis, and by the mould Aspergillus nidulans mentioned earlier⁴⁰.

The substance most associated with “cordyceps” today is therefore well known from the more ordinary caterpillar fungus, the one that can be grown on rice. According to these results, the expensive Himalayan fungus has hardly any.

How much cordycepin is there in cultivated fruiting bodies of the scarlet caterpillarclub? More recent analyses give something of the order of 1 to 4 milligrams per gram of dry weight, but published values range from roughly 0.05 to 18 mg/g, depending on strain, growth medium, light and method^{49, 50}. A figure without the strain and the method therefore says little.

Light plays a large part. In the dark the scarlet caterpillarclub does not turn orange, and Korean researchers name light as the single most important factor in cultivation⁵¹. In the strain studied, light affected the activity of 1,722 genes⁵².

Growers know one more question: why does a strain that has fruited for a long time start to degenerate? Under repeated subculturing, fruiting in the experiments of Sung’s group stayed stable up to the second to sixth subculture and then dropped sharply⁵³. In some of the strains studied, loss or a marked imbalance of mating types contributed to the degeneration^{54, 55}. Sensitivity, however, depends on the strain: in one study, production strains tolerated mating-type ratios from 1:9 to 9:1 without loss of yield, and a strain with a single mating type fruited up to the 11th generation, only without sexual spores (ascospores)⁵⁵. Forming fruiting bodies and completing sexual reproduction are therefore not the same thing. Strain degeneration is also accompanied by changes in metabolism and in how the cell reads its genes⁵⁶. For culture collections, the lesson is caution: limit the number of serial transfers and check for each culture that it still fruits. A universal recipe, such as an ideal mating-type ratio or the one correct way of freezing, does not follow from these studies.

Since 2018 it has been possible to edit the genes of the scarlet caterpillarclub in a targeted way using CRISPR⁵⁷. And cordycepin, meanwhile, has become independent of the fungus. Engineered yeasts and moulds produce it in grams per litre of nutrient solution; the mould Aspergillus oryzae, known from soy sauce production, does so in four days^{58, 59}. The best strains of the scarlet caterpillarclub are still ahead in quantity, but they take 45 to 60 days⁵⁹. What remains for the fungus is mainly what it has always been: a living organism that changes and needs to be watched.

↑ back to contents

09Cultivation: three things under one word

Today the words “cultivated cordyceps” can mean three entirely different things, and labels have historically lumped them together^{2, 62}.

The first are fermented mycelia. From the 1970s and 1980s, mycelium isolated from specimens of the Chinese caterpillar fungus was grown in fermentation tanks in China, then dried and ground for capsules; this gave rise to medicines such as Jinshuibao and Bailing^{2, 63}. Which fungi were really growing in the tanks was uncertain from the start. In 2015 Chinese mycologists wrote that the lack of a classification system for medicinal cordyceps fungi was a source of confusion⁶⁴. The most famous of those strains, Cs-4, turned out to be a different species altogether, Samsoniella hepiali⁷. It reached the West as CordyMax Cs-4, and in 2001 the US National Library of Medicine entered it in its index as a mycelial fermentation product of Cordyceps sinensis⁶⁵.

The second are genuine fruiting bodies of the Chinese caterpillar fungus, grown on reared moth caterpillars. For a long time this barely worked: in nature the caterpillars take three to five years to develop, and under artificial conditions the fungus did not form fruiting bodies^{66, 48}. Since the mid-2010s Chinese producers have been reporting tonnes. Caihong Dong’s group gives 2.5 tonnes in 2014, 5 tonnes in 2015 and 10 tonnes in 2016⁶⁷, another company 15 tonnes in 2017⁶⁶. These are the producers’ own figures, and no independent verification was found. A figure from the same review shows how hard this is: of 148 caterpillars that died in one rearing facility, 122 had been infected by another fungus⁶⁶.

The third are fruiting bodies of the scarlet caterpillarclub grown on grain, and these are exactly what MycoMuse Factory grows too. Their laboratory history begins in Europe and America. Anton de Bary grew the fungus in artificial culture in 1867, but without mature perithecia, the tiny flask-shaped structures in which spores form inside asci²; mature perithecia in laboratory culture were described by L. Shanor in 1936⁶⁸. Mass cultivation, however, is East Asian. A Korean book by J.-M. Sung from 1996 and Chinese papers from the first years of the 21st century describe cultivation on brown rice⁶⁸. According to Chinese reviews, China was the first country to master large-scale production of fruiting bodies on the pupae of the silkworm and of a giant silkmoth⁶².

In 2006 J.-M. Sung’s group at Kangwon National University published a method that can be summed up in one sentence. Two mutually compatible strains, each grown from a single spore, are inoculated into brown rice with silkworm pupae and left to grow at 20 °C, high humidity and continuous light, and 40 days later there are fruiting bodies⁵³. In a 2012 variant it was 50 grams of brown rice, 10 grams of pupae and 70 millilitres of water in a one-litre bottle, with the harvest after 60 days⁶⁸. Vietnamese growers work with rice at 24 °C and twelve hours of light a day, and harvest after 50 to 70 days². According to a 2024 review, the worst enemy of this kind of cultivation is “white mould”, the fungus Calcarisporium cordycipiticola⁶⁹.

When China approved the scarlet caterpillarclub as a new resource food in 2009, the fungus became an industrial crop. In 2019 products made from it had an annual value of 10 billion yuan in China, and 86 health foods containing the scarlet caterpillarclub are registered on the state platform⁷⁰. One Chinese review puts production in 2018 at more than 90,000 tonnes, but it is the only source, and it does not make clear whether this is fresh or dry weight⁷¹.

MycoMuse Factory grows it on the same principle, only on a small scale: in transparent containers of brown rice and nutrient solution under lights. Notes are kept on what was inoculated when, and how it grew.

Where people pay by the gram, there will also be those who add grams. In 1996 Taiwanese doctors described four cases of lead poisoning from caterpillar fungi into which someone had inserted small lead rods; once ground into powder, these could not be detected⁷². In February 2016 the Chinese drug authority warned that powders and tablets made from the Chinese caterpillar fungus contained four to ten times more arsenic than the limit allows⁷³. A Thai survey of 65 products bought online showed that more than half used the scarlet caterpillarclub in place of the Chinese caterpillar fungus, and four contained undeclared corticosteroids⁷⁴. In 2024 a US programme against the adulteration of botanical raw materials described supplements consisting mainly of grain growth medium with a minimum of mycelium, and wild caterpillar fungi weighted with bits of wire or made outright from flour and dye⁷⁵.

In the scientific literature, substitutions are detected with three tests: species by DNA, cordycepin and adenosine by HPLC, and heavy metals. Without them nothing certain can be said about what a product contains. The MycoMuse Factory lab page therefore states that an analysis of cordycepin and adenosine by HPLC would require a collaboration⁷⁶.

↑ back to contents

10Gold rush on the plateau

For a long time the Chinese caterpillar fungus was collected as a sideline. Under the people’s communes (1956–1981) there were state quotas, typically three pieces per person per day; during the Cultural Revolution the market collapsed, and in the 1970s a kilogram cost 21 yuan in Xining³. Then China opened up and the price began to rise. A kilogram at wholesale in Lhasa cost 1,800 yuan in 1985, 8,400 in 1997 and 36,000 in 2004³. At the end of 2006 the best pieces sold in China’s port cities, in Hong Kong and in San Francisco for 240,000 yuan per kilogram¹⁰, and in Beijing in 2017 for more than 140,000 dollars per kilogram, that is, for more than three times the price of gold⁷⁸.

For Tibetan and Himalayan herders this was a gold rush in the literal sense. In 2004 the fungus accounted for about 40% of rural cash income in the Tibet Autonomous Region, and in the places where it grows often for 70–90% of families’ annual cash income³. In the Nepalese mountain communities examined by a 2014 study, it is the second-largest source of household income after farming: 21% of all income and 53% of cash income, most of all among the poorest, and the money goes on school fees, food and paying off debts⁷⁹. In the Nepalese district of Dolpa alone, about 50,000 people took part in the harvest in the 2010 season⁸⁰. So every spring tens of thousands of people head into the mountains and spend two to three weeks searching the grass for a tiny dark stalk on which next year’s school fees hang.

Outside Asia, though, it was apparently athletics, not medicine, that made it most famous. In September 1993, at the National Games in Beijing, Chinese women runners set world records over 1,500, 3,000 and 10,000 metres, and Wang Junxia improved the 10,000 metres record by almost 42 seconds⁸¹. Their coach, Ma Junren, attributed this to altitude training and to tonics made, among other things, from caterpillar fungus and turtles^{82, 83}. The world marvelled, and the caterpillar fungus got publicity that no one could have bought. Later events complicated the story. In 2000 China dropped all but one of Ma’s runners from its Olympic team, officially because of injuries, form and positive internal tests⁸⁴. In 2016 a Chinese news website published a letter from 1995, signed by ten athletes including Wang Junxia, according to which the coach had run systematic doping. The International Association of Athletics Federations announced that it would verify the letter’s authenticity, and the Reuters agency could not confirm it independently^{85, 86}. The 1993 records are therefore not evidence of an effect of the fungus, whatever each of those performances may have been.

Meanwhile the harvest is shrinking. In Dolpa the average collector gathered about 261 pieces per season in 2006, and only 126 in 2010, despite spending more days in the mountains; 94% of the pieces were pulled out before they could release their spores⁸⁰. A 2018 study in the journal PNAS attributes the decline to overharvesting and warming: the fungus does better in colder places near permafrost, and winter temperatures there have risen in places by as much as 4 °C since 1979^{87, 78}. Nobody knows exactly how much of the fungus is collected each year. Estimates hover around 80 to 185 tonnes and have hardly changed in fifteen years. One explanation suggests itself: they are passed on from source to source rather than measured again^{10, 62}.

In July 2020 the IUCN listed the Chinese caterpillar fungus as a vulnerable species; according to the assessment, it had declined by at least 30% over the previous 15 years. Gregory Mueller of the IUCN called it at the time one of the few documented cases of a fungus threatened by overharvesting^{88, 89}. In September 2021 China listed it as a Class II protected species⁹⁰. Bhutan permits collection only in a limited season, India’s Uttarakhand only with a licence and outside reserves and national parks^{91, 92}, and in 2023 Nepal was preparing similar rules⁹³.

A simple line of reasoning follows from this. The scarlet caterpillarclub is common in Europe, is not threatened according to expert assessment, and can be grown in a transparent container. Interest in “cordyceps” therefore need not burden a threatened Himalayan species, though only if the label honestly says which cordyceps is in the packet.

↑ back to contents

11What the regulators say

The regulators agree on one thing: they distinguish between species, between products and between ways of use. Beyond that, their decisions diverge. This overview describes the situation as traced up to October 2026 and is not legal advice.

The European Union. A food that people in the EU did not eat to a significant degree before 15 May 1997 is a “novel food” and may be placed on the market only with an authorisation⁹⁴. The Dutch Food and Consumer Product Safety Authority (NVWA) states explicitly in its 2025 inspection report that the scarlet caterpillarclub is an unauthorised novel food in the EU and that supplements containing it therefore cannot be sold; when it inspected 17 companies, it found products with the scarlet caterpillarclub at five of them, and these companies had to withdraw them from the market⁹⁵. The EU Novel Food Catalogue says the same: the record for the scarlet caterpillarclub, covering both mycelium and fruiting bodies, lists it as a novel food that requires authorisation under Regulation (EU) 2015/2283 before it is placed on the market; the catalogue gives the record’s last update as 18 September 2025⁹⁶. The specialist press also reported on this classification during 2025^{94, 9}. According to the European Commission, the catalogue is an orientation aid, not a binding or complete list⁹⁷.

The same catalogue lists the Chinese caterpillar fungus under the name Ophiocordyceps sinensis, with the synonym Cordyceps sinensis, as a food that is not novel, but only in food supplements; other uses could be considered novel and require authorisation^{98, 95}. Its names in ten languages include the Czech housenice čínská⁹⁸. The catalogue gives the same status, “not novel in food supplements”, to Samsoniella hepiali (formerly Paecilomyces hepiali), the species to which the Cs-4 strain belongs^{99, 9}. The Belgian authority permitted Cordyceps sinensis in supplements as early as 2001, because it found that its consumption in Europe before 1997 had been significant¹⁰⁰. But the status of one fungus does not apply to another, and the name under which a product is sold does not in itself mean that the name is correct.

That is why fruiting bodies from MycoMuse Factory are not given out for eating or as a food supplement, not even free of charge¹⁰¹. Under the regulation on the general principles of food law, placing food on the market also includes transferring it free of charge¹⁰².

Czechia. No specific position of the Czech authorities on caterpillar fungi was found. Supplements containing Cs-4 mycelium are sold in pharmacies⁸. Red lists of fungi do not yet have legal force in Czechia³⁶.

China makes different distinctions. The then Ministry of Health approved the scarlet caterpillarclub as a new resource food by Announcement No. 3 of 2009; Announcement No. 10 of 30 May 2014 then changed its permitted amount, quality parameters and scope of use^{103, 70}. For the wild Chinese caterpillar fungus, the authority launched a pilot programme for its use in health foods in 2012. By a notice of 26 February 2016 it ended the programme and made products containing the fungus subject to the ordinary rules for registering health foods; without approval they may be neither produced nor sold¹⁰⁴. In the same month the authority warned about arsenic in powdered products made from the Chinese caterpillar fungus; the notice ending the programme does not itself mention arsenic^{73, 104}. Since 2021 the Chinese caterpillar fungus has been a protected species in China⁹⁰.

The United States. Caterpillar fungi are sold there as dietary supplements, and a trade handbook allows both the scarlet caterpillarclub and Cs-4 to be called “cordyceps”⁷⁵. The claim that cordyceps is officially recognised as safe in the US (GRAS status) appears in one chapter of a specialist book without a citation; no such notice was found in the database of the US FDA^{2, 105}.

↑ back to contents

12What is known about the effects

Now the chapter that many people read first, and which is placed almost at the end on purpose: how caterpillar fungi affect human health.

The short answer: the studies so far do not provide a reliable basis for general health promises about caterpillar fungi. Some preparations changed laboratory markers or selected performance outcomes; what that means for health, and whether it also holds for other preparations, remains uncertain. The longer answer is worth it, because it shows how a test-tube experiment becomes a promise on a packet.

In the test tube. Hundreds of papers have been published on cordycepin. A systematic review from 2021 went through more than a thousand of them and included 791. In experiments on cells grown in the laboratory, isolated cordycepin suppressed their survival and multiplication, their movement and inflammatory signals¹⁰⁹. These experiments did not test what this means for humans.

In animals. The same review found 150 animal studies, and its authors write that cordycepin has clear beneficial effects in many animals¹⁰⁹. Whether these results carry over to humans has not been verified. Drug development is further complicated by the fact that cordycepin is rapidly broken down in the blood by the enzyme adenosine deaminase; a pharmacokinetic study described this in rats as early as 2010^{110, 42}.

In humans: sport. A 2025 meta-analysis summarised four small studies with preparations labelled as Cordyceps sinensis, about 96 people in total, and one study with scarlet caterpillarclub mycelium in 22 runners. All were randomised and controlled. The result for endurance was at the very borderline of statistical significance, and the authors themselves describe it that way; for four of the five studies they found concerns about the risk of bias¹¹¹. A 2026 review devoted only to the scarlet caterpillarclub found five studies with 321 people aged 16 to 35, with doses of 1 to 12 grams a day for 1 to 16 weeks. Some recorded improvements in selected measures of performance or recovery, but the results were not consistent, and according to the authors its use in athletes should for now be regarded as “investigational rather than evidence-based”¹⁰⁶.

In humans: kidneys. Kidney disease is the subject of the largest number of studies. All the studies included in two large reviews came from China, and most of them tested medicines made from fermented mycelium, not fruiting bodies. A Cochrane review from 2014 covered 22 studies with 1,746 adults with chronic kidney disease, who were given the preparations in addition to standard treatment. Some laboratory markers improved, but because of the low quality of the studies no definitive conclusions could be drawn¹¹². An umbrella review from 2024 summarised 157 studies in diabetic kidney disease. It rates the methodological quality of 13 of the 14 reviews it covers as critically low and the certainty of the results as very low to moderate, and no included study reported mortality or kidney failure¹¹³.

In humans: immunity. In a Korean study from 2015, 39 healthy men received capsules with the scarlet caterpillarclub for four weeks and 40 men received a placebo; in the group given the fungus, NK cell activity and some other markers of cell-mediated immunity increased¹¹⁴. A Thai study from 2024 gave 40 healthy adults either a drink made from scarlet caterpillarclub mycelium powder or a placebo, daily for eight weeks; some immune markers in the blood changed, others did not, and the study has small groups and inconsistencies in reporting⁶⁰. A change in a blood marker does not by itself mean less illness.

Aphrodisiac. For the oldest traditional use, the one in the Tibetan text, the literature search found no study in humans.

Tumours. Intravenous cordycepin and its synthetic successor NUC-7738 are cancer drugs under investigation, not a fungus in food (see the chapter on Glasgow)^{42, 45}.

Safety. The genome of the scarlet caterpillarclub contains no genes for known human mycotoxins⁴⁶, but that does not prove that the products are safe. The documented problems mainly concern adulteration and contamination: lead, arsenic, undeclared drugs and allergies to silkworm pupae in the growth medium^{72, 73, 74, 62}.

Why can none of this be carried over? Many studies “with Cordyceps sinensis” actually used fermented mycelium of other fungi, and almost none measured cordycepin. A result with mycelium in a capsule is not a result with a fruiting body, and an intravenous infusion of a pure substance is not tea. History (the Tibetan recipe, the duck from Qing cookery, the 1993 records) has the value of a story, not of evidence.

For now, science’s answer here is: this is not yet known. It is an honest answer, and there is no need to be ashamed of it.

↑ back to contents

13Caterpillar fungi in Czechia and at MycoMuse Factory

MycoMuse Factory is a small grow room and budding lab in the Šumava. The fungus grown here is the scarlet caterpillarclub, not the Chinese caterpillar fungus. In a tent there is shelving, on it transparent containers, in the containers brown rice with nutrient solution, above them white and red light that is on for fourteen hours a day, and at each container a camera that takes a picture every ten minutes while the light is on; alongside, sensors record the temperature, humidity and CO₂ in the tent⁷⁶. The method follows Korean and Chinese cultivation from the turn of the millennium, only on a much smaller scale.

For a grow room this small, three things above all follow from the literature.

A strain can weaken with repeated subculturing. Growers have known this for decades, and in some strains it is linked to the loss or imbalance of mating types. That is why the strain code, batch and passage are recorded for each culture⁷⁶. Verifying the mating types of the starting cultures by PCR is, for now, only on the list of things a collaboration would help with⁷⁶.

A number without a method means nothing. The cordycepin content of fruiting bodies from MycoMuse Factory is not given for now. An analysis would require HPLC, because ordinary UV spectrometry at 260 nm cannot tell cordycepin from adenosine, and this analysis too is on the list for collaboration⁷⁶. On differences between strains, the lab page states that with one or two containers per strain these are pilot observations, not a proven difference⁷⁶.

What is not written down cannot be verified years later. That is why raw images are not altered, data are only added, and a correction is a new record with a reference to the original⁷⁶. The data are not public for now⁷⁶; the website shows the course of cultivation: images, stages and a diary.

And one more thing: the fruiting bodies from MycoMuse Factory are not for eating, even when they are given away free; the chapter on the regulators explains why. Containers with living fruiting bodies are given out only for observation¹⁰¹.

↑ back to contents