This study was conducted using roxatidine acetate or Mylanta combined with metronidazole and amoxicillin to evaluate the role of acid and Helicobacter pylori in the natural course of the duodenal ulcer. Eighty-three patients with H. pylori positive duodenal ulcers were randomly allocated into one of four treatment groups. Group A: roxatidine 75 mg hs for 8 weeks; Group B: the same as group A + metronidazole 250 mg and amoxicillin 250 mg qid for 1 week on the 3rd week; Group C: Mylanta (combined hydroxide of magnesium and aluminum) 20 ml qid for 8 weeks; Group D: the same as group C + metronidazole 250 mg and amoxicillin 250 mg qid for 1 week on the 3rd week. Repeated endoscopies were performed on the 8th week post the initial treatment and the sixth and 12th month post the termination of treatment, or, at the earliest recurrence of symptoms. Eradication of H. pylori was considered to be successful if the culture, histology and CLO test all showed negative. The ulcer healing rates of Groups A, B, C and D were 95%, 100%, 61% and 86%, respectively, with a significant difference between A and C. The eradication rates of groups B and D were 81% and 62%, respectively, without any significant difference. The 12 months cumulative ulcer recurrence rates were 72%, 15%, 80% and 22%, respectively, with a significant difference between each paired groups with and without antibiotics. In conclusions, roxatidine is effective in the healing of duodenal ulcer. One-week roxatidine-based triple therapy is powerful in the eradication of H. pylori. Potent acid suppression is sufficient to heal the duodenal ulcer. Eradication of H. pylori can potentiate ulcer healing under insufficient acid suppression. A causal role for H. pylori in recurrent duodenal ulcer is strongly supported by a much lower recurrence of ulcer in H. pylori free patients.