Ninety-eight patients with erosive and/or ulcerative esophagitis unhealed after at least 3 months' treatment with standard doses of cimetidine (greater than or equal to 1200 mg daily) or ranitidine (greater than or equal to 300 mg daily) were primarily included in an acute healing phase study, and 51 were allocated to 40 mg omeprazole once daily and 47 to 300 mg ranitidine twice daily. After 12 weeks of treatment, 46 (90%) patients given omeprazole were healed, compared with 22 (47%) allocated to ranitidine. Healed patients were then given maintenance treatment with either 20 mg omeprazole once daily or 150 mg ranitidine twice daily for 12 months. Plasma gastrin was determined and gastric mucosal biopsy specimens were obtained during the entire study to assess the structure of the exocrine and endocrine cell populations of the oxyntic mucosa. Sixty-seven per cent of the total number of patients randomized to omeprazole were maintained in clinical and endoscopic remission throughout the 12-month study period as compared with only 10% among those given ranitidine (p less than 0.0001). After 4 weeks of omeprazole treatment basal gastrin levels were slightly increased, with a 95% confidence interval for the change of from 8.6 to 16.9 pmol/l. No further increase in basal gastrin levels was observed during the ensuing study months. No significant histopathologic lesion was found in the oxyntic gland mucosa. In conclusion, omeprazole was far superior to ranitidine in preventing recurrence, a goal achieved without adverse events and significant abnormalities in the oxyntic mucosal exocrine or endocrine cells but with a moderate increase in basal gastrin levels.