Thirty-six consecutive patients, who were to be treated with cisplatin-based chemotherapy for testicular or bladder cancer, underwent a single-blind randomized study to compare the antiemetic therapies with dexamethasone (DEX)+ondansetron (OND) and DEX + alizapride (ALI). Eighteen patients were assigned to each arm. DEX, 20 mg in 100 ml saline was administered i.v. 30 min prior to cisplatin, OND, 8 mg, or ALI, 100 mg in 100 ml saline were administered i.v. 15 min prior to cisplatin and repeated 4 and eventually 8 h later. Chemotherapy regimens contained cisplatin 25 mg/m2 for 4 consecutive days to be repeated for 4 courses every 4 weeks. During the first course a complete emetic control was observed in 15 (83%) and partial in 3 of the 18 patients treated with DEX + OND versus only 2 complete and 7 partial responses and 9 (50%) failures among the 18 patients treated with DEX + ALI. Thirty-one patients were evaluable for 4 courses of therapy. Complete emetic control was achieved in 11 (69%) and partial in 5 (31%) among the 16 patients treated with DEX + OND, versus only 1 (7%) partial response and 14 (93%) failures among the 15 treated with DEX + ALI (p < 0.001). Furthermore, DEX + OND gave a complete antiemetic control in 13 out of 14 patients who had failed DEX + ALI. Delayed vomiting was observed in 4 (22%) of 18 patients primarily treated with DEX + OND and in 1 (7%) of the 15 patients subsequently treated. Constipation and headache occurred more frequently among patients treated with DEX + OND, but there was no significant difference with DEX + ALI. Hiccup was significantly more frequent among patients treated with DEX + ALI.(ABSTRACT TRUNCATED AT 250 WORDS)