Ulcer healing and relapse rates after treatment with cimetidine and sucralfate in 112 subjects with endoscopically proven gastric and duodenal ulcers were compared in a randomized trial. Four grams of sucralfate were administered in four divided doses each day; 200 mg cimetidine was administered three times during the day with the fourth dose of 400 mg given at night. The subjects were divided into four treatment groups: (1) subjects with duodenal ulcers treated with sucralfate; (2) subjects with duodenal ulcers treated with cimetidine; (3) subjects with gastric ulcers treated with sucralfate; and (4) subjects with gastric ulcers treated with cimetidine. Endoscopy was repeated at 6 weeks after treatment was initiated; subjects with unhealed ulcers at 6 weeks were reendoscoped at 12 weeks. Eighty-three percent of duodenal ulcers in subjects treated with sucralfate healed at 6 weeks, the rest by week 12. Cimetidine-treated subjects with duodenal ulcer had healing rates of 71 and 86% at 6 and 12 weeks. Cimetidine-treated subjects with gastric ulcers showed healing rates of 75 and 89% at 6 and 12 weeks, respectively. In the gastric ulcer group, 63% of sucralfate-treated subjects were healed at 6 weeks, and 78% were healed by 12 weeks, respectively. Eighty-six of the subjects were followed for 1 year or until endoscopy showed relapse. Subjects treated with cimetidine on a short-term basis relapsed earlier than those were treated with sucralfate at 12 weeks, but at 1 year, the relapse rates were about 70% for both groups.