Background: Rabeprazole has a faster onset of antisecretory action than omeprazole, and it is of interest to determine whether this translates into faster symptom relief in patients with gastro-oesophageal reflux disease. Aims: To assess the relief from heartburn after 3 clays of treatment with standard-dose rabeprazole or high-close omeprazole (primary end-point). Secondary end-points included the decrease in score for other symptoms of gastro-oesophageal reflux disease, healing rates and quantification of antacid use. Methods: Patients with endoscopically confirmed erosive oesophagitis were randomized to receive 4 weeks of double-blind treatment with rabeprazole (20 mg) or omeprazole (40 mg). Patients who were not healed after 4 weeks received a further 4 weeks of treatment. Results: Two hundred and seventy-four patients were screened, 251 patients were randomized and 230 patients completed the trial. The numbers of patients with relief from heartburn on day 4 were similar in the two groups (84% for rabeprazole; 95% confidence interval, 76-90%; 83% for omeprazole; 95% confidence interval, 75-89%). There were no significant differences between the treatments in the relief from other gastro-oesophageal reflux disease symptoms or in healing rates. The number of reports of severe heartburn during the first 3 days was higher in the omeprazole group (daytime heartburn: 4.7% for rabeprazole vs. 10.3% for omeprazole, P = 0.005; night-time heartburn: 4.7% for rabeprazole vs. 9.8% for omeprazole, P = 0.01; statistical comparisons defined post hoc). Conclusions: Standard-dose rabeprazole was as effective as high-close omeprazole in relieving symptoms by day 4 of treatment and in healing oesophageal lesions, but had a faster onset of action in patients with severe heartburn. This suggests that the improved pharmacological properties of rabeprazole translate into a clinically relevant advantage.