Aspirin is commonly accepted as a risk factor for gastric ulcer; however, there is little published evidence linking aspirin consumption to duodenal ulcer. The effect of 1 g of aspirin per day on site-specific ulcer hospitalizations was examined using data from a 3-year randomized, double-blind, placebo-controlled trial of 4,524 subjects (Aspirin Myocardial Infarction Study). There were 23 duodenal ulcer and 14 gastric ulcer hospitalizations during the follow-up period. All but two were verified by endoscopy, radiogram, or biopsy/surgery. For males, a Cox-model survival analysis showed that the age- and smoking-adjusted relative risk for duodenal ulcer hospitalization was 10.7 times higher for the aspirin group than for the placebo group (95% confidence interval, 2.5 to 45.5; p less than 0.0001). The adjusted relative risk for gastric ulcer was 9.1 (95% confidence interval, 1.2 to 71.4; p = 0.04). Due to the small number of females in the study, the relationship between site-specific ulcer and aspirin consumption for females was not analyzed. However, for males and females combined, the age-, smoking-, and sex-adjusted relative risk for peptic ulcer hospitalization was 7.7 (95% confidence interval, 2.7 to 21.7; p less than 0.0001). We conclude that chronic aspirin use is a risk factor for hospitalization for both duodenal and gastric ulcer in males, and for peptic ulcer in males and females.