Eighty patients with Campylobacter pylori-associated duodenal ulcer disease were randomly allocated to receive colloidal bismuth subcitrate (CBS)tablet 120 mg four times a day or ranitidine 150 mg twice daily in a trial comparing the effects of these drugs in short-term healing and post-healing relapse rates of duodenal ulceration. At 8 weeks 88.1% (37/42) of those on CBS and 92.1% (35/38) of those on ranitidine had ulcers healed. The difference is not significant. After ulcer healing, the cumulative rates of relapse, as determined endoscopically, for symptomatic and asymptomatic ulcers were 19.4% (6/31) for CBS and 46.7% (14/30) for ranitidine at 6 months (P less than 0.05) and 41.9% (13/31) for CBS and 73.3% (22/30) for ranitidine at 12 months (P less than 0.05). As campylobacter pylori was cleared in 35 of the 42 patients (83.3%) in the CBS group, while only one of the 38 patients (2.63%) in the ranitidine group (P less than 0.005), it is possible that the clearance of Campylobacter pylori by CBS is instrumental to the reduction of the rate of reulceration.