The course of gastritis and Campylobacter pylori was studied in a single-blind randomized trial comparing cimetidine 200 mg three times a day and 400 mg at night and sucralfate 1 g four times a day orally for four weeks in 140 patients with proved duodenal ulcer. At least two antral biopsies were performed during endoscopy before entry and at the end of four weeks. The activity and the degree of chronic inflammation, as assessed histologically by the degree of infiltration of, respectively, polymorphs and chronic inflammatory cells, were graded blindly by two pathologists as nil, mild, moderate, or severe. The density of C. pylori, as assessed after Warthin-Starry stain, was similarly graded. Ulcer-healing rates were comparable in the cimetidine (73.2 percent) and sucralfate (79.7 percent) groups. Improvement of the activity of gastritis occurred significantly (p less than 0.05) more frequently in the sucralfate (33.3 percent) than in the cimetidine group (18.3 percent), and remained so (p less than 0.05) when only patients with healed ulcer were compared. The density of C. pylori decreased significantly in the sucralfate group after treatment (p less than 0.01) but not in the cimetidine group. The 12-month ulcer relapse rates were significantly (p less than 0.05) lower by life-table analysis in patients healed with sucralfate than in those healed with cimetidine and were unaffected by either the density of Campylobacter in either group or the improvement of the gastritis. It is concluded that sucralfate improves duodenal ulcer-associated antral gastritis and decreases the density of C. pylori, and that factors other than bacterial density and antral gastritis may be responsible for the advantage of sucralfate over cimetidine in ulcer relapse.