The Stomach That Proved Medicine Wrong: How Barry Marshall Drank Bacteria and Won a Nobel Prize

On a Tuesday morning in July 1984, a 32-year-old doctor at Fremantle Hospital in Australia asked his lab technician to prepare him a drink. The technician scraped the contents of two petri dishes full of bacteria into a beaker of beef broth. Marshall swallowed the murky brown liquid quickly, like a shot. He did not tell his hospital superiors or his wife. He just went home and waited to get sick.

Within days, his breath turned foul. His wife and his mother both noticed before he admitted anything. Then came the bloating, the nausea, the vomiting every morning. Ten days after drinking the bacteria, an endoscopy showed his stomach lining was inflamed and full of the corkscrew-shaped organism he had swallowed. He had successfully given himself gastritis, which was exactly what he had been trying to do. He then took a course of antibiotics, cured himself, and began writing up the results.

What the Medical World Believed Before

For most of the 20th century, the standard explanation for stomach ulcers was stress. The theory made intuitive sense. Ulcers seemed to flare during periods of emotional strain. They were associated with high-pressure lifestyles and anxiety. The treatment matched the theory: antacid medications to reduce stomach acid, dietary changes, stress management, and in severe cases surgery to cut the nerves that stimulated acid production.

This was not fringe medicine. It was the consensus. It was what doctors worldwide learned in medical school and what they taught their patients. The antacid market was enormous and enormously profitable, one of the most successful drug categories in the pharmaceutical industry. The entire framework, treatment, diagnosis, research priorities, was built on the premise that ulcers were a stress-related condition managed but never cured.

Nobody was looking for a bacterium because the idea seemed absurd. The stomach produces hydrochloric acid strong enough to dissolve metal. Nothing lives in there. That was simply known.

How Robin Warren Started the Whole Thing

The story starts not with Marshall but with a pathologist named Robin Warren, who was working at Royal Perth Hospital in Western Australia in 1979.

Warren was examining biopsy samples from patients with gastritis and stomach ulcers, and he kept seeing something he wasn’t supposed to see. Small, curved, spiral-shaped bacteria, living directly in the stomach lining. Every textbook said the stomach was sterile. Warren was looking at slides that said otherwise.

He reported what he saw. The medical community mostly dismissed it as contamination. Tissue samples sometimes picked up bacteria after the fact, and the assumption was that this was happening here. Warren was looking at dead samples, the logic went, not live infection.

Warren disagreed. He noticed that wherever the bacteria appeared, the surrounding tissue showed signs of inflammation. Where no bacteria appeared, the tissue was healthy. He kept seeing this pattern. He could not unsee it. In 1981, he convinced a young internal medicine registrar named Barry Marshall to take a look.

Marshall Takes It Seriously

Marshall reviewed Warren’s slides and found himself convinced almost immediately. The correlation was too consistent. He started culturing the bacteria from patient biopsies, a technically difficult process that took months of failed attempts before succeeding. He and Warren ran a study of 100 patients, examining biopsies and correlating bacterial presence with disease. The results were striking: nearly every patient with gastritis or peptic ulcer had the bacteria.

They submitted their findings to conferences and journals. The response ranged from skepticism to ridicule. One conference rejected their abstract. Reviewers questioned their methodology. The consensus position, that ulcers were caused by stress and acid, had decades of research and billions of dollars of pharmaceutical investment behind it. Two researchers in Western Australia saying bacteria were actually the cause was not a welcome finding.

Additionally, they couldn’t infect animal models. Rats and mice didn’t get ulcers from H. pylori. Without animal models, they couldn’t satisfy Koch’s postulates, the standard requirements for proving a pathogen causes a disease.

The Beaker

This is where Marshall ran out of patience. He had already treated eleven patients with difficult duodenal ulcers using bismuth and an antibiotic. Nine of them got better. Without the bacteria, their ulcers didn’t return. The treatment was working. Nobody wanted to hear it.

So in July 1984, Marshall cultured bacteria from a patient with gastritis. He confirmed the patient’s infection was present. Then, without telling anyone, he mixed the culture into a beaker of broth and drank it. His reasoning was straightforward. If the bacteria caused gastritis in a healthy person, that proved the causation chain. It wasn’t just correlation. The bacteria came first, then the disease. Koch’s postulates could be satisfied using himself as the subject.

It worked faster than he expected. Within days he had symptoms. Within ten days, his biopsy confirmed active gastritis and H. pylori infection. He had been healthy before. He was infected now. The bacteria had done what he said they would do.

Why It Took So Long to Change Anything

Marshall published the self-experiment results. The medical establishment remained largely unmoved. This is the part of the story that deserves more attention than it usually gets. Marshall had now satisfied the causation requirement. He had given himself the disease and cured it with antibiotics. The evidence was there. And still, mainstream gastroenterology took another decade to fully accept it.

Part of this was institutional inertia. The stress-acid model had enormous investment behind it. Pharmaceutical companies had profitable antacid products. Hospital protocols were built around long-term acid management. Changing all of that required not just scientific evidence but organizational will, regulatory movement, and the slow replacement of one generation of practitioners with another.

Part of it was also legitimate scientific caution. One self-experiment on one doctor in Australia was not, by the standards of evidence-based medicine, sufficient to overturn decades of established practice. Marshall’s findings needed replication, larger trials, and peer review across multiple independent research groups.

That process happened, but slowly. In 1994, the National Institutes of Health formally declared that H. pylori was the primary cause of most stomach ulcers, recommending antibiotic treatment as the standard of care. This was ten years after Marshall drank the beaker.