Background: Droperidol and 5-HT3-receptor antagonists are among the most potent antiemetics to prevent postoperative nausea and vomiting (PONV). Combinations of these drugs have been used to increase the efficacy of antiemetic treatment. However, so far the quantitative effect of this combination has not been evaluated systematically. Methods: Results from randomised controlled trials investigating the efficacy of 5-HT3-receptor antagonists or droperidol alone versus the combination of both drugs to prevent PONV were included in a meta-analysis. Studies were systematically searched using Medline, EMBASE, the Cochrane-Library, and by manually screening the reference lists of matching review articles and current issues of locally available peer-reviewed anaesthesia journals. Seven papers with data on granisetron published by Fujii and co-workers were not considered. The main end point in each study was defined as occurrence of nausea, retching, or vomiting within 6 h (“early PONV”) and within 48 h (“late PONV”) after surgery. The relative risks (RR) and the numbers needed to treat (NNT) of the pooled data with their corresponding 95% confidence intervals (given in parentheses) were calculated using a random effects model. Results: Eight studies with 881 patients (adults: n=801; children (mean age: 8 yr): n=80) were included in the analysis. Droperidol was applied to 340 patients, 5-HT3-receptor antagonists to 198, and 343 were treated with a combination of both drugs. Seven out of these eight studies reported increased antiemetic efficacy of the combination group compared with the single drugs (droperidol and 5-HT3-receptor antagonists respectively). However, in none of the trials did this difference reach statistical significance. When a meta-analytic analysis based on these results was performed the combination of droperidol with a 5-HT3-recpetor antagonist was not associated with a significantly increased antiemetic efficacy. In 12 to 13 patients a 5-HT3-receptor antagonist has to be added to droperidol prophylaxis to prevent one additional patient from PONV who would have had suffered from PONV when treated with droperidol alone (RR “early PONV”: 1.52 (0.95–2.44); RR “late PONV”: 1.24 (0.89–1.74)). Similar results were obtained when the antiemetic effect of adding droperidol to a prophylaxis with 5-HT3-receptor antagonists was analysed. In this case 10 to 12 patients have to be treated with the 5-HT3-droperidol combination instead of with a 5-HT3-receptor antagonist alone to prevent one additional patient from PONV (RR “early PONV”: 1.55 (0.68–3.52); RR “late PONV”: 1.29 (0.77–2.17)). There were no reports of an increased incidence of adverse effects. Conclusion: The data on the combination of droperidol with 5-HT3-receptor antagonists suggest that there is a trend towards increased efficacy of the combination therapy compared to the single drugs. However, so far there are insufficient data to recommend this combination treatment for prophylaxis.