Listen to this essay
28 minute listen
In late summer 2023, I found myself in a smoke shop looking at a package of ‘magic mushroom’ gummies.
That phrase, magic mushrooms, usually calls to mind one type of mushroom: those containing psilocybin, the compound showing promise as a therapy for depression, addiction and other psychopathologies. But these gummies were not psilocybin. Instead, the packaging described Amanita muscaria, the red-capped, white-spotted toadstool of fairy-tale forests, iPhone keyboards and Super Mario.
I knew a little about the folklore of Amanita muscaria. I had first seen it discussed in an episode of the show Hamilton’s Pharmacopeia (2016-21), which begins with a claim I now hear often: that Amanita muscaria may be among the oldest human intoxicants. The New York banker-turned-ethnomycologist R Gordon Wasson, whose Life magazine essay from 1957 introduced psilocybin mushrooms to a mass Western audience, certainly took this idea seriously. In his book Soma: Divine Mushroom of Immortality (1968), he argued that the Vedic soma – a ritual drink in ancient Indian religions – was made from Amanita muscaria. The mushroom has also been linked, sometimes carefully and sometimes wildly, to the origins of Santa Claus, Viking berserkers and even Christianity. Some believe its other name, ‘fly agaric’, stems from its use as a natural pesticide for house flies; the word muscaria is also derived from the Latin word musca, meaning fly. Some scholars have linked Amanita muscaria to Lewis Carroll’s mushroom scene in Alice’s Adventures in Wonderland (1865), in which the Caterpillar tells Alice that ‘one side’ of the mushroom will make her taller and the other shorter, though Carroll does not identify a species.
But I wasn’t standing in that smoke shop holding one of these products as a folklorist or looking to lose myself in Carroll’s ‘golden afternoon’. I am a public-health scientist, and my questions were pragmatic: what exactly was in this package? Could it hurt someone? Were these products legal?
The first thing I did was what I usually do. I went to the scientific literature. And down the rabbit hole I tumbled.
Although manufacturers use the term ‘magic mushrooms’, the main psychoactive compound in Amanita muscaria is not pharmacologically interchangeable with psilocybin; the two drugs do not act on the brain in the same way.
The body converts psilocybin into psilocin, the substance that actually produces its psychedelic effects. Psilocin switches on serotonin 5-HT2A receptors, proteins on the surface of brain cells that respond to the chemical messenger serotonin. These receptors are concentrated in the cortex, the brain’s outer layer, and especially in the prefrontal cortex, the region behind the forehead involved in such functions as attention, perception and our sense of self. The receptors sit on the big output neurons that broadcast to the rest of the brain, and switching them on makes those cells fire harder. Brain scans of people on psilocybin show what follows: networks that normally run in step with each other come apart, and regions of the cortex that rarely signal to each other begin to. What people report from the inside is a change in perception rather than false perception – surfaces that appear to move, colours that intensify, time that stops running at a fixed rate.
Muscimol killed mice at a lower dose than nearly all psychotropic drugs we examined, including fentanyl
Amanita muscaria works through a different system. Its principal psychoactive compound, muscimol, acts primarily on GABA-A receptors – part of the brain’s main braking system – quieting neuronal activity rather than amping it up. That puts muscimol broadly on the inhibitory side of brain chemistry, the same signalling system influenced by central nervous system depressants such as benzodiazepines and alcohol. The experience is still ‘psychoactive’, but it can look quite different. Clinical reports of Amanita intoxication describe drowsiness and dream-like states, impaired balance and coordination, confusion and disorientation, sometimes alternating with agitation, as well as visual or auditory hallucinations. In severe cases, people can become profoundly sedated, delirious or lose consciousness. In other words, rather than the characteristic alterations in perception associated with psilocybin, Amanita can produce a much less orderly mixture of sedation, hallucination and confusion.
The risks are different too. When my colleagues and I began looking at the toxicology, the numbers startled me. By the crude measure toxicologists often begin with – the LD50, or the dose required to kill half the test population – published estimates suggested that muscimol administered orally killed mice at a lower dose than nearly all commonly used psychotropic drugs we examined, including fentanyl. As a vegetarian, I found the exercise grim, and translating animal doses to humans is always imperfect. But these were safety signals we took seriously.
That was the first turn in the rabbit hole. The next was dosing. If companies were selling Amanita muscaria gummies and chocolates in ordinary retail settings, then somewhere, implicitly or explicitly, someone had made a decision about dosing. A gummy contains some amount. A package contains some total amount. A label tells a consumer to take one piece, or half a piece, or perhaps more. A certificate of analysis from products submitted to a laboratory can provide an amount per gummy. Those numbers can feel reassuring simply because they are printed on a package. They also imply that someone knew enough to choose a dose for the desired effect.
I wanted to know whether that chain existed.
So, I started with a simple question: are there human studies showing what muscimol does at different doses? The best answer to this would have come from controlled studies in which multiple human participants were given standardised doses under observation, with careful reporting of effects and adverse events. I could not find that literature in the usual places. I could find toxicology estimates, case reports, historical accounts and ethnobotanical discussions. But a modern dosing study, the kind one would want before placing psychoactive candies near a cash register, was not there.
I kept looking.
Eventually, I came across the proceedings from a symposium I had not known about: the Ethnopharmacologic Search for Psychoactive Drugs, held at the University of California, San Francisco, from 28 to 30 January 1967. The time and place stood out to me immediately. Two weeks earlier, a couple of miles away in Golden Gate Park, Timothy Leary had told the Human Be-In crowd to ‘turn on, tune in, drop out’, a phrase that would become shorthand for the moral panic that soon surrounded psychedelics and particularly LSD. The symposium also opened with a letter by Albert Hofmann, the Swiss chemist best known for synthesising LSD and who had also led the work identifying psilocybin and psilocin as the active compounds of Psilocybe mexicana. The symposium covered compounds from many psychoactive plants and cultures, including a lecture on nutmeg delivered by Alexander Shulgin – the San Francisco Bay Area biochemist who is remembered most infamously (and perhaps unkindly) for his rediscovery and popularisation of MDMA (aka Ecstasy).
A mid-century scientist, dosing himself, may have produced numbers that later travelled into a marketplace of gummies
Near the end came a session on Amanita muscaria. In the final part of the final lecture, I found what I had been searching for – something resembling a dosing guide. It came from a report of a self-experiment with pure muscimol by Peter G Waser – a Swiss pharmacologist who later served as rector of the University of Zurich – under observation by a Swiss psychiatrist named Jules Angst. Waser repeatedly dosed himself with pure muscimol dissolved in water to establish a range of effects. This was not a modern dosing study. It was stranger than that, and I think it is worth hearing Waser’s language directly because it captures both the scientific seriousness and the unease of the moment:
Curiosity is one of the main qualities of a scientist, and this may be the reason why a pharmacologist watching the behaviour of his animals under certain drugs wants to know more about the emotions and changes in reaction they produce. Another reason is ethical. We have to foresee the accidents and dangerous actions new compounds of general interest may produce on man. My report on some experiments done on myself by ingesting ibotenic acid and muscimol under careful psychiatric control may give you an unauthoritative view on this.
At 5 milligrams taken orally in water, Waser reported little effect beyond a feeling of laziness. At 10 milligrams, however, he described a mild intoxication beginning about 90 minutes later, with dizziness, loss of coordination and elevated mood. He did not experience hallucinations, although he noticed slight changes in taste and colour vision, followed by muscle twitching and then sleep with dreams. Within two to three hours, he reported feeling normal again – rested and able to work – and said he slept deeply the following night.
At 15 milligrams, the effects were stronger and began more quickly, after about 40 minutes. Waser reported dizziness severe enough that he could not walk with his eyes closed, along with slurred speech, diminished appetite and taste, and increasing difficulty concentrating. His visual perception also changed: images of situations he had seen minutes earlier seemed to repeat themselves, and sounds sometimes appeared to echo. Most disturbing to him were repeated involuntary muscle contractions. At times, he wrote, he felt as though he had lost his legs, although he specifically distinguished the experience from the vivid, colourful hallucinations he had experienced with LSD. After falling asleep and awakening several hours later, he felt dull and uncertain rather than refreshed, although the muscle twitching had diminished. His blood pressure rose only slightly during the psychoactive period. In preliminary analyses conducted by the Swiss chemist Conrad H Eugster, muscimol appeared to be detectable in Waser’s urine after the self-experiment, although Waser cautioned, at the time, that the finding still needed confirmation.
What I found was upsetting, but not that surprising. Dosing recommendations from product manufacturers, including the one I found at that smoke shop, appeared to resemble that old self-experiment. While I have no way to confirm it, I have a strange feeling that this mid-century scientist, dosing himself under observation and reporting on it at a famous drug conference, may have produced numbers that later travelled into a marketplace of gummies and chocolates sold to consumers who likely have no idea where the guidance came from.
I wish I had been able to speak with Waser himself, but he died in 2010. In 2024, I did exchange emails with Angst, the psychiatrist who had supervised the experiment. Angst was 97 and not in good health, but he confirmed the point that mattered most to me: Waser had come to Angst for supervision as a precaution. Waser’s own later writing suggests why. Recalling the experiments, Waser described the experiences at high doses as unpleasant and even potentially life-threatening.
Convinced of questionable dosing standards in this marketplace, my line of enquiry returned to the potential harms – were there products on the market that contained enough muscimol to seriously harm someone?
I found a product with a very high reported amount of muscimol. It was a gummy product, one in a benign-looking package that did not have childproofing or clear warnings. I began doing the calculation no one wants to do. If a young child ate the entire package, mistaking it for candy, could they die? I took the reported amount of muscimol, the toxicology estimates (the LD50), a conversion used to translate dosing between species, and the average weight of a small child. The result looked bad. Bad enough that I did not quite trust myself. So, I called a friend, Matthieu Rouffet, a medicinal chemist and the dean of science at a local college, who does toxicology calculations more often than I do. I wanted him to tell me I had made an error. Instead, he was alarmed too.
There is a rare moment as a scientist when you encounter an empirical finding and it becomes emotionally charged. Before, it might just have been intellectual curiosity and your training that drove you to it. After, you feel a sense of responsibility. This was one of those rare moments for me. Perhaps this finding was especially emotionally charged for my friend Matthieu and me because we are fathers of young children. When I made the calculation, my daughter was only a year old. It could have just been the sleep-deprived delirium of early fatherhood, but I felt a moral responsibility to act.
The broader pattern appeared to be permissiveness by omission or legal loophole
Were there policy levers I could pull? It was complex because Amanita muscaria fell into the gap between the categories – not a controlled substance, not a food, not quite a supplement.
Unlike psilocybin, Amanita muscaria and muscimol were not scheduled under the US Controlled Substances Act, and they also did not appear to be controlled under comparable drug laws in most countries, with a few exceptions. The Netherlands treated Amanita muscaria and Amanita pantherina (which also contains muscimol) as controlled substances, and, in the US, manufacturers generally acknowledged Louisiana as a state where sales were restricted. One manufacturer’s legal guidance listed prohibitions in countries including Brazil, Thailand, Australia, Romania, Russia, Sweden and Ukraine. But the broader pattern appeared to be permissiveness by omission or legal loophole.
This was an argument I had seen a few years earlier, when I wrote about the ‘hemp loophole’. It emerged from the 2018 Farm Bill legalising hemp, defined largely by its low level of delta-9-THC, the main intoxicant in cannabis. Some manufacturers saw an opening: if the law restricted just delta-9-THC, they argued, other intoxicating cannabinoids derived from legal hemp could still be sold. One was delta-8-THC, often produced from hemp-derived CBD, a nonintoxicating cannabinoid. Soon delta-8 gummies, vapes and candies were turning up in gas stations, smoke shops and online stores. The industry’s argument was essentially this: if the starting material was legal hemp and the finished product stayed below the federal limit for delta-9-THC, it did not have to be sold through the regulated marijuana market, with safeguards such as age restrictions and testing for potency.
The Amanita muscaria market seemed to be borrowing a similar playbook. The claim was not that Amanita muscaria gummies had been proven safe. It was that they were not psilocybin, not scheduled, and not clearly prohibited.
However, drug scheduling was only one possible policy lever. Food and dietary supplement law was another. Many Amanita muscaria products were being sold as if they were ordinary foods or supplements, but I could not find publicly available notifications showing that Amanita muscaria or muscimol had been accepted as a new dietary ingredient in the US or elsewhere.
Regulators could clarify whether these products could be commercially manufactured and sold at all
Given the toxicology and case reports mentioned above, that absence mattered. The issue, here, was not whether someone could pick a mushroom in the woods and take it home to eat – that’s unambiguously legal. The issue was whether a company could extract, concentrate, flavour, package, advertise, and sell a psychoactive compound in candy-like forms without meeting basic consumer protection standards.
They didn’t appear to be meeting this standard, and so there was legal recourse. Regulators could clarify whether these products could be commercially manufactured and sold at all under dietary supplement and food laws. If sales were not allowed, they could take the products off the shelves. If sales were allowed, they could require product standards such as age restrictions, child-resistant packaging, maximum-dose rules, warning labels, restrictions on cartoons or candy-like marketing, and standardised testing for potency, contaminants and undisclosed drugs. They could also require labels to state what was in the package in units that were interpretable.
Now, with a clear safety concern and a clear pathway for legal recourse, my colleagues and I wrote a paper, published in a peer-reviewed journal, and we began raising awareness on the subject. I issued a press release with my school at the University of California, San Diego, which helped to generate some national media coverage including in The Washington Post and on NPR. I also contacted colleagues at the Department of Justice in my home state of California. They put me in touch with the consumer-safety division that would eventually build an investigation of several brands selling Amanita muscaria-branded products in the state, including one that led to the seizure of more than 95,000 mushroom chocolate bars from a warehouse in San Diego County.
By late 2024, the US Food and Drug Administration (FDA) had also reached its own conclusion about Amanita muscaria in food, offering a perspective that confirmed my own. In a scientific memorandum, FDA toxicologists concluded that Amanita muscaria and three of its active constituents, including muscimol, do not meet the criteria for safe use in food in the US.
There’s another side to the story. Some people believe the mushroom helps them. There are books and online communities organised around microdosing Amanita muscaria, often motivated by purported benefits based on, and following dosing information derived from, survey responses or anecdotal reports. The health claims tend to cluster around sleep, anxiety and the possibility of reducing dependence on benzodiazepines or other sedatives. Muscimol’s pharmacology makes some of this plausible in a broad mechanistic sense. A compound acting primarily on the GABA-A receptors could, in theory, have sedating effects that are therapeutic. Even our forerunner Waser alludes to this in his self-experiment. That account also mentions the findings of his colleague Eugster, who held a patent on a method of synthesising muscimol, which is currently assigned to the pharmaceutical company Novartis. Less toxic analogues of muscimol have also been explored as pharmaceuticals; gaboxadol, for instance, was discovered by the Danish chemist Povl Krogsgaard-Larsen. But the benefits and harms of these substances have never been rigorously tested to this day.
A thousand testimonials about the use of Amanita muscaria do not tell you what is in a gummy from a particular batch made by a particular company. They do not establish a safe range for children, older adults, pregnant people, people taking other medications, or anyone who eats more than the label recommends because the product looks harmless.
Once a substance is turned into a consumer good, science and capitalism are already in the room
This is one of the tensions running through the modern conversation about psychedelics and plant medicines like Amanita muscaria. Some proponents of plant medicine argue that science should not be treated as the only legitimate way of evaluating the veracity of therapeutic claims. They point to traditional and Indigenous uses of psychoactive plants and fungi, sometimes over millennia, and ask why randomised controlled trials should outrank lived experience, ritual knowledge or community practice.
I take that argument seriously. Western medicine does not own all wisdom, and therapeutic approval pathways are limited. Many plant medicines will struggle through conventional drug approval because they cannot be easily patented, because no company has an incentive to fund the studies, or because they belong in a ceremonial setting that is not transferable to a randomised trial. If the only knowledge we recognise is knowledge that can be monetised through drug development, we will miss a great deal.
But the Amanita muscaria products sold in smoke shops are not simply traditional plant medicines being used in traditional contexts. They are commercial products in a modern marketplace. They are manufactured, packaged, branded, advertised, and shipped. They borrow the language of science, from microdosing to certificates of analysis, when it helps them. Once a substance is turned into a consumer good, science and capitalism are already in the room. The question is whether they will be used with integrity.
The distinction between plant medicine and a commercial market became painfully clear with the recall of Diamond Shruumz-brand chocolate bars, ice-cream cones and gummies sold by Prophet Premium Blends, based in Santa Ana, California. In 2024, federal agencies investigated 180 cases, including three potential deaths, tied to products marketed with mushroom language and sold in the same broad retail ecosystem of gas stations, smoke shops, hemp-derived cannabinoid stores and online sellers.
In their recall notice, Prophet Premium Blends indicated that the adverse events may have been due to elevated levels of muscimol. But product testing complicated that story. Muscimol did not appear in all products and could not explain all the reported symptoms. Instead, testing found a shifting mixture of substances across products, including psilacetin, a near chemical cousin of psilocin first disclosed in a patent by Albert Hofmann and Franz Troxler in 1961. Unlike psilocin and psilocybin, psilacetin is not always listed by name in controlled-substance schedules, though it may still be regulated under analogue laws or broader drug-control provisions. Some samples also contained psilocin, pregabalin (a prescription medication), and compounds related to the psychoactive root, kava. The result was not a simple story about muscimol toxicity. It was a more troubling story about a marketplace where consumers could not reliably know what psychoactive substances they were consuming.
That, to me, illustrates the central ethical issue I see in the conversation about modern psychedelics and plant medicines.
People can argue about whether adults should be allowed to use psychedelics or plant medicines. My own view is that adults should generally have wide latitude over what they put in their own bodies. But they also have a right to know what they are putting in their bodies, as well as the known benefits and risks. A person cannot meaningfully consent to ingesting one thing while being sold another. They cannot weigh risks that are hidden from them inside of mislabelled gummies. They cannot dose carefully when the dose is uncertain and the active ingredient is undisclosed. The right to alter one’s consciousness does not include a company’s right to deceive.
Psilocybin was creating a halo around other mushrooms, and manufacturers were exploiting it
And while there are many stories about the purported benefits of plant medicines like Amanita muscaria, there are also stories about the harms of this marketplace. The most impactful story I’ve learned about came from Ashley Glover, who contacted me through email in 2024.
Ashley first wrote to me after reading my peer-reviewed article. Her only brother Andrew Davenport died on 23 June 2024, she said, after consuming a Diamond Shruumz mushroom bar. What stood out in Ashley’s emails was not only the grief, but the difficulty of getting the justice system to recognise what might have happened. She wrote that a witness who was with Andrew had insisted that the bars made him sick. She wrote that an opened box of Diamond Shruumz bars was recovered near his body. She also wrote that his initial toxicology report did not show hard drugs or alcohol, and that his prescribed medications were described to the family as being at normal therapeutic levels. But the medical examiner, she said, had not saved the box or bars for testing and the explanation relayed to her was blunt: no one overdoses on magic mushrooms.
That sentence has stayed with me because it captures the potential danger of a category error. If the product had been treated as a psilocybin mushroom, its risk could be dismissed before the relevant chemistry was even known. But these products were not psilocybin mushrooms. They were manufactured psychoactive goods whose contents could vary, whose labels could be incomplete, and whose compounds were difficult for forensic scientists to detect.
Ashley told me that biological samples from Andrew were eventually sent for specialised testing. Whether those results can establish causation is a legal and forensic question, and I do not want to claim more than the evidence can support. But her account shows the public-health problem in human terms. Once a consumer product enters the marketplace without reliable safety evidence, accurate labelling and adequate testing capacity, the uncertainty does not end at the point of sale. It follows the consumer, the family, the medical examiner, the investigator, and anyone trying to reconstruct what happened after an adverse event leaves questions in its wake.
I began this investigation with a question that seemed narrow enough for an academic paper – should unregulated sales of Amanita muscaria products prompt a public-health response? The answer, to me, became an unequivocal yes. But the journey was stranger than the paper could contain.
I did not set out to become the public-health scientist warning people about the fairy-tale mushroom, sounding alarmist or like a teetotaller. I am not opposed to research on psychoactive substances or to responsible use among adults. If anything, the therapeutic promise of psilocybin is part of why Amanita muscaria bothered me; psilocybin was creating a halo around other mushrooms, and manufacturers were exploiting it in ways that I saw as unethical and dangerous.
The lesson I took from Amanita muscaria was not that altered states are inherently dangerous or that science should flatten every human encounter with plants or fungus into a risk calculation. It was simpler than that. If a company sells a psychoactive product to the public, especially in forms that look like candy, it should be transparent about what is in the product, and have evidence that the product can be used without unreasonable harm. This should not be a radical standard. It should be the minimum requirement for sales.