This multicentre, double-blind, double-dummy, randomised trial was designed to compare the efficacy and safety of single intravenous doses of dolasetron mesilate and granisetron in the prevention of acute emesis and nausea due to high-dose (> or = 80 mg/m2) cisplatin. Single intravenous doses of 1.8 or 2.4 mg/kg of dolasetron mesilate or 3 mg of granisetron hydrochloride were administered in a volume of 50 ml over a 5-min period, beginning 30 min prior to cisplatin (> or = 80 mg/m2) administration. The number and timing of emetic episodes, time to administration of escape anti-emetic medication, severity of nausea by visual analogue scale (VAS), and safety were monitored for 24 h after the start of cisplatin-containing chemotherapy. Investigators' evaluations of overall efficacy and patients' satisfaction with therapy were recorded at the end of the 24-h study period. Of the 474 patients evaluable for efficacy, complete responses were achieved by 54, 47 and 48% of patients given dolasetron mesilate 1.8 mg/kg, dolasetron mesilate 2.4 mg/kg and granisetron, respectively. Statistically, treatment groups had comparable complete and complete plus major responses, times to first emesis, and use of escape medication; patient maximum nausea severity and treatment satisfaction ratings; and physician nausea severity and overall efficacy assessments. For the majority of efficacy endpoints, 1.8 mg/kg dolasetron mesilate produced numerically superior responses compared with the 2.4 mg/kg dose. Gender and prior chemotherapy were significant predictors of complete response; males and chemotherapy-naive patients had higher responses. The overall incidences of adverse events were comparable among the treatment groups; headache and diarrhoea were most common. In conclusion, 1.8 and 2.4 mg/kg of dolasetron mesilate and granisetron (3 mg) were equally effective in preventing nausea and vomiting induced by highly emetogenic cisplatin-containing chemotherapy. In addition, because no additional benefit was observed with 2.4 mg/kg of dolasetron mesilate and numerically greater responses were observed with the 1.8 mg/kg dose, the lower dose of 1.8 mg/kg is optimal for further clinical development.