We performed a systematic review of the effectiveness of anti-emetics for prophylaxis of cisplatin-induced delayed emesis using meta-analysis. We selected 12 reports of randomized controlled trials from MEDLINE (1966-2003. 4) and The Cochrane Library Issue 1, 2003. Nine of these reports were evaluated as high quality and the others as low quality according to the evaluation criteria of Jadad et al., and only the high-quality reports were subjected to meta-analysis. The statistical results obtained from all 12 reports were also compared with those obtained from the 9 reports of high quality. Corticosteroids significantly reduced the occurrence of delayed emesis. Metoclopramide tended to reduce the occurrence of delayed emesis, although not to a significant extent. In contrast, 5-HT3 receptor antagonists did not show a significant prophylactic effect on delayed emesis. Combination treatments using corticosteroids with metoclopramide or 5-HT3 receptor antagonists did not show significant additional benefits over corticosteroids alone. In conclusion, treatment with corticosteroids without additional metoclopramide or 5-HT3 receptor antagonists appears to be preferable for the prevention of delayed emesis induced by cisplatin.