In a double-blind randomized trial 91 patients with endoscopically confirmed duodenal and/or prepyloric ulcers received either 1000 mg cimetidine daily (44 patients) or placebo (47 patients) for 4 weeks. At 2 weeks complete ulcer healing was significantly increased in patients receiving cimetidine (45%) in comparison with those treated with placebo (15%) (p less than 0.01). At 4 weeks 73% of the patients on cimetidine and 32% of those on placebo showed complete healing (p less than 0.001). Ulcer size decreased more rapidly with cimetidine than placebo (p less than 0.05). With respect to the number of pain-free days and nights cimetidine was significantly superior to placebo during the days of the second week only (p less than 0.05). Cimetidine patients consumed significantly less antacids than those receiving placebo (p less than 0.01). There was no significant correlations between ulcer healing and freedom from pain at the end of either cimetidine or placebo treatment in this study. No specific side effects referable to cimetidine were observed.