This study was designed to compare the rates of duodenal ulcer healing and recurrence after treatment with cimetidine or antacid. Patients with endoscopically documented duodenal ulcer received cimetidine, 1200 mg daily, or Mylanta II, 7 oz daily, in a randomized, double-blind trial. For the 69 patients in each group who completed the healing phase of the trial, endoscopic ulcer healing was almost identical. At 2, 4, and 6 wk, the cumulative percent healed on antacid was 33%, 64%, and 80%, and on cimetidine it was 25%, 62%, and 86%. The 114 patients with healed ulcer were observed on no therapy and underwent additional endoscopy to detect recurrences when symptomatic or at 3, 6, and 12 mo. There was no difference in the frequency of recurrences between treatments. At 3 and 6 mo, the cumulative percentages of patients with recurrence were 29% and 56% after antacid therapy and 36% and 55% after cimetidine therapy. Some patient variables were associated with delayed ulcer healing or ulcer recurrence. These included sex, pain frequency, smoking, disease duration, and acid secretion.