His neighbour turned into a malevolent witch that afternoon. At least, that is how she looked to the chemist blinking at her from his own doorway: thirty-seven years old, mild-mannered, and deep into a compound he had invented five years earlier and then written off as a dud.

Nobody at his company had wanted it.

A fungus with a long criminal record

Ergot infects rye. For centuries it poisoned entire villages, and the sufferers of ergotism got burning limbs, convulsions and visions that medieval Europe filed under St Anthony’s fire. Sandoz, the pharmaceutical firm in Basel, looked at the same fungus and saw a chemistry set. Its alkaloids narrow blood vessels, which makes them useful for stopping bleeding after childbirth and for treating migraine.

Albert Hofmann was hired to take those alkaloids apart, strip them back to the core they all share, lysergic acid, then attach new fragments to it and see what happened. He joined Sandoz in 1929 with a doctorate from Zurich. By the late 1930s he was grinding through the series one molecule at a time.

What the lab was actually chasing

Not a psychedelic. Sandoz wanted an analeptic, meaning a drug that props up breathing and circulation, and the template was Coramine, a product made by Ciba on the opposite bank of the Rhine. Coramine is a diethylamide, so Hofmann bolted a diethylamide group onto lysergic acid to see whether the same trick would transfer. He finished the twenty-fifth compound in the series on 16 November 1938, a date recorded in Mystic Chemist, the biography by Dieter Hagenbach and Lucius Werthmüller. Hence LSD-25.

Animal testing that year was a flop. Pharmacologists logged strong activity on the uterus, roughly seventy per cent as potent as ergonovine, a drug Sandoz already sold. The only other note in the file, preserved in a history published by the Multidisciplinary Association for Psychedelic Studies, was that the test animals grew restless under anaesthesia. Nothing there looked like a product. Ergot was expensive, war was coming, and the batch was set aside.

Into the cupboard it went.

Five years, then a hunch

Shelved compounds at Sandoz stayed shelved. Hofmann later admitted he could never shake the feeling that the compound had been dropped too early. That is an odd thing for a methodical man to admit, and odder still that he acted on it. On Friday 16 April 1943 he made a fresh batch.

Mid-afternoon he stopped. Dizzy and unable to concentrate, he went home, lay down and spent the next two hours watching colours and shapes drift behind his closed eyes. He suspected the new compound, absorbed in traces through his fingertips. Rolling Stone, drawing on his journals, later put the accidental dose at around 20 micrograms.

Monday afternoon, 4.20

Confirming a suspicion like that meant drinking some. Hofmann dissolved 250 micrograms in water and swallowed it at twenty past four on 19 April, a quantity he thought conservative because ergot alkaloids normally need milligram doses to do anything at all. LSD is active at a fraction of that. Forty minutes later his notebook entries fell apart into fragments, and he asked his laboratory assistant to take him home. Petrol was rationed, so they went by bicycle.

That ride is the reason 19 April now has a name. In LSD: My Problem Child, Hofmann describes the world warping as though seen in a curved mirror, and a sensation of being rooted to the spot no matter how hard he pedalled. His assistant told him afterwards that they had travelled at a cracking pace.

A doctor came, found enormous pupils and an otherwise unremarkable patient, and left. Terror gradually gave way to something else. Sounds turned into shapes, the witch became a neighbour again, and Hofmann slept. Speaking decades later at a conference, as NPR has recounted, he said the heightened sensitivity lasted into the following day.

The dud gets a second career

Sandoz decided it had a psychiatric tool on its hands, marketed it under the name Delysid and made it widely available to researchers, a chapter reconstructed in detail by Chemistry World. More than a thousand papers followed. Then the compound walked out of the clinic, joined a decade it is still associated with, and was banned almost everywhere.

Regulators are now circling back. A dose-finding trial of a pharmaceutical LSD formulation in 198 adults with generalised anxiety disorder, published in JAMA in October 2025, reported a dose-dependent drop in anxiety scores after a single supervised administration, sustained at twelve weeks. That is one industry-funded trial with a company on the byline, and two phase 3 studies in the same condition are now running to test whether it holds.

Hofmann died in 2008 at the age of 102, as reported at the time by NBC News, still maintaining, as he put it for decades, that it was the compound that had found him, not the other way round. Making it took an afternoon in 1938. Working out what it was for has taken everyone else the better part of a century.