OBJECTIVES: To compare the efficacy of esomeprazole with the European licensed standard dose of PPIs for the endoscopic healing of reflux oesophagitis (i.e. esomeprazole 40mg compared with lansoprazole 30mg, omeprazole 20mg, pantoprazole 40mg, and rabeprazole 20mg). METHODS: Systematic review of CENTRAL, BIOSIS, EMBASE and MEDLINE for randomised controlled trials (RCTs) in patients with reflux oesophagitis was conducted in February 2005. Data on endoscopically verified healing rates at four and eight weeks were extracted and re-analysed if not analysed by intention-to-treat. The summary effect estimate (Relative Risk, RR) was calculated by meta-analysis using a fixed effects model by the Mantel-Haenszel method. The quality of RCTs included in the meta-analysis was judged according to the method of randomisation and concealment of allocation of treatments (as this is the element of RCT design likely to introduce the most bias). Publication bias was assessed using funnel plots. RESULTS: Of the 133 papers initially identified in the literature search, only 8 were found to be head-to-head comparisons of a standard dose of PPI compared with esomeprazole 40mg. This total was reduced to six when the quality of the RCTs was assessed. From the 6 trials, a meta-analysis of endoscopic healing rates of esomeprazole 40mg compared with standard-dose PPIs gave the following results: at 4 weeks, RR 0.92 (95% CI: 0.90 to 0.94, p<0.00001); at 8 weeks, RR 0.95 (95% CI: 0.94 to 0.97, p<0.00001). The effect of using all eight trials in the meta-analysis made small numerical differences to the overall estimates but did not change the direction or make a significant difference non-significant. Publication bias did not appear to have a significant impact on the results, as there was no apparent asymmetry identified in the assessment of funnel plots. CONCLUSIONS: Esomeprazole demonstrates significantly higher endoscopic healing rates when compared to standard-dose PPIs.