Recent reports in the United States have found that fewer peptic ulcers are due to Helicobacter pylori than previously believed. The aim of this study is to determine if the declining prevalence of H. pylori infection in the general population can account for the apparent increase in the frequency of non-H. pylori ulcers. A total of 396 patients with peptic ulcer or ulcer scar were enrolled in this study. The pre-1950 population consisted of 149 patients with gastric ulcers and with 44 duodenal ulcers. The post-1950 population consisted of 96 patients with gastric ulcers and 107 with duodenal ulcers. The frequency of H. pylori-negative gastric ulcers was 5.4% in patients born before 1950 and 4.2% in patients born after 1950, and the frequency of H. pylori-negative duodenal ulcers was 0% and 1.9%, respectively. There are no statistical differences between the two populations in gastric and duodenal ulcers. H. pylori seropositivity was 74.9% in asymptomatic volunteers born before 1950 and 20.7% in those born after 1950 (P < 0.01) in the general population. The attributable risk of H. pylori infection in peptic ulcer diseases was not affected by the prevalence of H. pylori infection in the general population in Japan. This suggests that the apparent increase in frequency of non-H. pylori ulcers in the United States is not simply due to the declining prevalence of infection. Other explanations for non-H. pylori ulcers should be sought.