Fifty-eight patients with endoscopically confirmed benign gastric ulceration were randomly allocated to treatment with 150 mg ranitidine twice daily, placebo matching ranitidine twice daily, or 200 mg cimetidine three times daily and 400 mg at night. Patients were endoscoped at monthly intervals for up to 3 months, the endoscopist being unaware of the treatment. Significantly more ulcers (p less than 0.05) had healed after 2 months of ranitidine (14 of 18, 78%) and cimetidine (17 of 20, 85%) than with placebo (9 of 20, 45%; p less than 0.05) and after 3 months of ranitidine (15 of 18, 88%) and cimetidine (18 of 20, 90%) than with placebo (11 of 20, 55%; p less than 0.05). Forty-eight patients with healed ulcers were randomly allocated in a double-blind prophylactic study to receive 150 mg ranitidine at night or matching placebo. After 6 months recurrent ulcers were found in 2 of 24 (8%) of patients receiving ranitidine and 10 of 24 (42%) of patients receiving placebo (p less than 0.05). These data indicate that H2-receptor antagonists are significantly better than placebo in healing gastric ulceration and that ranitidine and cimetidine are equally effective. Ranitidine is significantly superior to placebo in preventing gastric ulcer recurrence.