A randomised controlled multicentre trial was performed in 160 patients with gastric ulcer, proved by endoscopy and biopsy, to compare ulcer healing with sucralfate and ranitidine (double blind double dummy design) and to assess the effect of maintenance treatment with sucralfate on ulcer recurrence (double blind placebo controlled design). The healing rates were similar with 4 g sucralfate suspension per day and 300 mg ranitidine per day (82% and 88% after 12 weeks, respectively). Of the 109 patients with healed ulcers, 92 were entered into the maintenance trial and treated with sucralfate tablets (2 g per day) or placebo tablets. Maintenance treatment with sucralfate delayed symptoms of gastric ulcer recurrence. Lifetable analysis showed significant differences between sucralfate and placebo, both after six months (p = 0.018) and after 12 months (p = 0.044). The rates of symptom recurrences were 13% and 34% after six months and 34% and 55% after 12 months for sucralfate and placebo, respectively. The rate of asymptomatic recurrences after 12 months was similar in the two groups (9% and 10%, respectively). The recurrence rate was higher in patients who had never taken non-steroidal anti-inflammatory drugs than in those who had but had stopped on admission to the study. It was also higher in patients with recurrent ulcer and in those with scarring deformation and narrowing of the pylorus. Maintenance treatment with sucralfate slowed the appearance of symptom recurrences of gastric ulcer.