BACKGROUND The efficacy of prokinetic agents in functional (nonulcer) dyspepsia has been questioned based on recent trial results. We performed a meta-analysis to determine the efficacy of cisapride and domperidone in functional dyspepsia. METHODS Computer and manual searching was used to identify placebo-controlled studies that included >20 patients. The statistical analysis focused on: global assessment by the investigator, epigastric pain, early satiety, abdominal distension and nausea (all rated on four-point scales). Results are reported as odds ratios (OR) in favor of treatment. Regression analysis was performed to evaluate possible effect modifiers. The relationship between improvement in gastric emptying and symptoms was also evaluated. RESULTS For cisapride, 17 studies met the inclusion criteria, but varying numbers of studies had to be used for the different outcome measures. For all outcome measures, there was a statiscally significant benefit in favor of cisapride: global assessment of improvement by the investigator or patient (OR 2.9, 95% CI 1.5-5.8), epigastric pain (OR 0.19, 95% CI 0.05-0.7), early satiety (OR 0.18, 95% CI 0.9-0.4), abdominal distension (OR 0.32, 95% CI 0.1-0.7), and nausea (OR 0.26, 95% CI 0.1-0.5). Age of patient, year of publication, and country where study was performed had only small modifying effects. There were insufficient data to determine whether there is a relationship between improvement in gastric emptying and response to treatment. For domperidone, four of eight studies could be used for the analysis of global assessment of improvement by the investigator. This showed an OR of 7.0 (95% CI 3.6-16) in favor of domperidone. CONCLUSIONS Both cisapride and domperidone seem to be efficacious in functional dyspepsia, although this conclusion is largely based on global assessment by the investigator, which may not be an optimal outcome measure.