OBJECTIVE The objective of this study was to compare the cost of achieving a unit of clinical success for treatment of reflux oesophagitis with either ranitidine or omeprazole. METHODS After randomisation 430 patients with reflux oesophagitis (grade 2 or 3) were assigned to receive omeprazole 20 mg or ranitidine 150 b.i.d. for 8 weeks. Patients were given diary cards to assess their symptoms every day, and record every two weeks a life satisfaction index. Patients were seen after 4 and 8 weeks for symptoms assessment and repeat endoscopy at 8 weeks. The perspective of the analysis was that of the payer. The costs of medical care were based in French drug costs currently advertised, payment for physician and actual mean payment for upper GI endoscopy. RESULTS The healing rates at 8 weeks in the omeprazole group and the ranitidine group were 93 and 67.5% respectively. After 8 weeks of treatment, life satisfaction was good in 81 and 53.6% and the relief of pain was 86,6 and 69,5% respectively. For each effectiveness criteria, omeprazole was more cost effective than ranitidine: cost per healed patient (2,338 F vs 2,744 F), cost per asymptomatic patient (2,510 F vs 2,964 F), cost per patient with a good or very good life satisfaction index (2,687 F vs 3,456 F). This advantage remained independent of the oesophagitis initial severity. The sensitivity analysis showed that the results were unsensitive to the variations of the efficacy variables in their confidence interval. CONCLUSIONS This cost-effectiveness analysis suggests that in the treatment of reflux oesophagitis, the strategy with the more effective treatment is more cost-effective.