The aims of this study were to assess the prevalence of duodenal ulcer during chronic pancreatitis and the relationship between the occurrence of duodenal ulcer and the course of chronic pancreatitis. The population consisted of 264 men with chronic pancreatitis, 94 percent of which were of alcoholic origin. Duodenal ulcer occurred in 37 patients (14 percent). Fifteen patients with gastric or anastomotic ulcer were excluded. The 37 patients with duodenal ulcer were compared with 212 patients without ulcer. There was no significant difference between the two groups as regards the course of chronic pancreatitis except for insulin-dependent diabetes mellitus which was significantly more frequent in patients without duodenal ulcer (P less than 0.05). Eight patients with duodenal ulcer died but the cause of death was not related to their ulcer. All patients with duodenal ulcer and 92 percent of those without were smokers (not significant). Duodenal ulcer occurred in 25/37 patients (68 percent) before the clinical onset of chronic pancreatitis. The risk of duodenal ulcer occurrence was constant in 17-65 year old patients and independent of the time of chronic pancreatitis onset. We concluded that in men with chronic pancreatitis: a) prevalence of duodenal ulcer is 14 percent; b) duodenal ulcer occurred most often before clinical onset of chronic pancreatitis; c) duodenal ulcer occurs independently of the course of pancreatitis which cannot account for its high prevalence. Smoking may be a promoting factor.