Campylobacter pylori infection has been associated with duodenal ulcer, gastric ulcer, and non-ulcer dyspepsia. Although in vitro studies have shown that C. pylori is susceptible to most commonly used antibiotics, predictions from in vitro sensitivity studies have not led to a safe and generally effective therapy; C. pylori has proved to be very difficult to eradicate in vivo. We used the urea breath test to assess the susceptibility of C. pylori in vivo to various drugs. C. pylori was susceptible to bismuth subsalicylate, bismuth subnitrate, and furazolidone. C. pylori was not susceptible (i.e., urease activity remained despite administration of the drug) to the following drugs: 1) antiulcer agents (cimetidine, ranitidine, famotidine, omeprazole, misoprostol, sucralfate, liquid antacids); 2) NSAIDs (aspirin, indomethacin, ibuprofen, naproxen, tolmetin); 3) antibiotics (oral penicillin V, trimethoprim-sulfamethoxazole, dicloxacillin); 4) salts (lithium, ferrous sulfate, gold); 5) miscellaneous (acetaminophen, phenytoin, hydrochlorothiazide, propranolol, metoprolol, metoclopramide, ursodeoxycholic acid). Oral antimicrobials can be administered directly onto the site of infection, so that a very low oral dose will provide many multiples of the in vitro minimal inhibitory concentration. Furazolidone suspension (7 mg) was administered seven times daily (daily dose 49 mg) to three individuals infected by C. pylori during suppression of gastric acid secretion with famotidine (40 mg bid). After 4 days, all subjects had significant reductions in urease activity (two to normal and one to a borderline value). This response suggested that very low-dose therapy may be useful either alone or combined with bismuth. Conclusive establishment of an etiologic (or major contributory) relationship of C. pylori to ulcer disease will require a safe and reliable method to eradicate the organism from the stomach and duodenum.