Three hundred seventy patients with recently healed duodenal ulcer entered into a one-year, double-blind, randomized multicenter trial comparing placebo with three different dose schedules of cimetidine (200 mg twice a day, 300 mg twice a day, and 400 mg at bedtime) for the prevention of recurrent duodenal ulcer. By the end of one year, the cumulative symptomatic recurrence rate as demonstrated by endoscopy was similar for the patients receiving the three dosages of cimetidine (19 per cent, 15 per cent, and 13 per cent, respectively; not significant), whereas the placebo-treated group had a 34.7 per cent symptomatic recurrence rate (P less than 0.01 as compared with each cimetidine group). Cigarette smoking was found to be an important variable; among the placebo recipients ulcer recurrence was significantly more likely in smokers (72 per cent) than in nonsmokers (21 per cent, P less than 0.001). The frequency of ulcer recurrence in smokers was significantly reduced by treatment with cimetidine (from 72 per cent to 34 per cent, P less than 0.). Smokers who received cimetidine were at least as likely to have a recurrence as were nonsmokers who received placebo (34 per cent vs. 21 per cent, not significant). Thus, smoking appears to be a major factor in recurrence of duodenal ulcer, and in smokers, giving up smoking may be more important in the prevention of ulcer recurrences than administration of cimetidine.