The aim of this study is to assess the relationship between Helicobacter pylori and the relapse of duodenal ulcer, and also to evaluate the differences in efficacy and side effects between 1 week and 2 week triple therapy. Sixty-two patients with active duodenal ulcer, which healed within 8 weeks of nizatidine treatment, were randomly allocated to one of two groups. Group 1 (n = 29) received no drugs, Group II (n = 33) received triple therapy for 1 week (IIa, n = 16) or 2 weeks (IIb, n = 17). Eleven patients whose ulcer did not heal after an 8 week nizatidine treatment period were randomly assigned into Group IIa (n = 5) and IIb (n = 6). Seven patients whose ulcer recurred after discontinuation of nizatidine were allocated to receive 2 weeks of triple therapy. All patients received endoscopy 6 weeks after entry, and again at 3, 6 and 12 months unless both ulcer recurrence and H. pylori infection were found. The frequency of ulcer relapse 6 weeks after the active duodenal ulcer had healed was 83% (24/29 in Group I, 13% in Group 11a and 14% in Group IIb. The cumulative rate of recurrence was significantly higher in Group I than in Group II (90 vs 30% at 12 months, P < 0.01). Ulcer relapse was associated with persistence of H. pylori infection (P < 0.0001).(ABSTRACT TRUNCATED AT 250 WORDS)