This paper is a meta-analysis of 30 published, double-blind clinical trials comparing omeprazole with ranitidine or cimetidine for the treatment of duodenal ulcer, gastric ulcer and reflux oesophagitis. These studies compare the recommended doses of omeprazole with those for ranitidine and cimetidine, and the confidence intervals for the therapeutic gain show that the findings are highly reliable. The difference in healing rates favoured omeprazole over ranitidine in patients with duodenal ulcer after 2 weeks of treatment (15.2 percentage units; P < 0.001), and after 4 weeks of treatment in patients with gastric ulcer (9.9 percentage units; P = 0.005), or reflux oesophagitis (23 percentage units; P < 0.001). Similarly, omeprazole gave a 20.6 percentage units higher average healing rate than cimetidine in patients with duodenal ulcer after 2 weeks of treatment (P < 0.0001). Significantly more patients treated with omeprazole were free of symptoms at their first follow-up visit than patients treated with ranitidine or cimetidine.