Abstract Eighty-five patients with endoscopically confirmed duodenal or pyloric canal ulcers entered a double blind trial with 1200 mg of cimetidine per day or placebo for 6 weeks. Eighty-four per cent of patients treated with cimetidine and 38% of those receiving placebo healed their ulcers ( P < 0.001). Measurement of basal acid output and maximal acid output before and after treatment showed no significant change but patients who failed to heal their ulcers had a higher basal acid output and maximal acid output than those who healed. Patients who smoked or drank alcohol had the same healing rate as abstainers. Sixty-seven patients with duodenal ulceration healed by a 6-week course of cimetidine were randomly allocated to 400 mg of cimetidine twice daily or placebo in the maintenance trial. Actuarial analysis of the number of relapses in each group demonstrates that cimetidine is highly effective in preventing relapse.