Forty three subjects with DU confirmed at endoscopy and healed after eight weeks of sucralfate therapy were subjected to a randomized double-blind controlled trial for six months. Twenty one subjects received sucralfate (1 g one hour before breakfast and dinner). Twenty two subjects received placebo. Symptoms were evaluated every month. The subjects were endoscoped at the end of the trial or earlier in case symptoms recurred. Sucralfate was found to be significantly more effective than placebo (6/21 vs. 17/22, p < 0.005) in preventing DU relapse. Age at presentation, age at onset, duration of illness, sex, periodicity, smoking, gastric stasis like symptoms, associated irritable bowel syndrome, degree of deformity of the bulb and initial presence of duodenitis were the factors examined for their effect upon the relapse. In the placebo group relapsers had significantly shorter mean duration of illness indicating that DU may relapse more frequently in earlier part of its natural course. Other factors did not influence the relapse rate. In the sucralfate group, higher ages at onset and at presentation were associated with significantly higher relapse rate. Sucralfate may be less effective in preventing relapse in elderly and late onset DU patients.