This randomised, double-blind, parallel-group study was carried out to compare the efficacy and safety profile of ondansetron plus dexamethasone and metoclopramide plus dexamethasone in patients receiving fractionated cisplatin (20-25 mg/m2/day) chemotherapy for the treatment of testicular cancer. An interim analysis of 95 patients showed that the ondansetron regimen was significantly superior compared to the metoclopramide regimen (p < 0.001). According to the study protocol the study was terminated at this stage. At the time the decision to stop the study was taken, a total of 113 patients had been enrolled and were evaluable on an 'intention to treat' basis. Fifty-six of these had received ondansetron (32 mg i.v. single dose/day) plus dexamethasone (20 mg i.v. single dose/day) and 57 were given metoclopramide (2 mg/kg or 1 mg/kg i.v. twice a day) plus dexamethasone (20 mg i.v. single dose/day). The ondansetron regimen was significantly superior in the control of emesis and nausea. Seventy-one percent of patients experienced 2 or fewer emetic episodes over the entire 5-day study period compared with 26% of patients given metoclopramide (p < 0.001). Seventy-nine percent of patients in the ondansetron group experienced 'none' or only 'mild' nausea compared with 39% of patients in the metoclopramide group (p < 0.001). The dose of metoclopramide had to be reduced during the study from 2 mg/kg i.v. twice daily to 1 mg/kg i.v. twice daily because 4 of the first 8 patients randomised to this treatment experienced extrapyramidal reactions. Ondansetron was well tolerated and it did not induce any extrapyramidal reactions. The results of this study show that ondansetron plus dexamethasone represents a very effective treatment option for patients receiving fractionated cisplatin chemotherapy for testicular cancer.