The objective of this study was to compare the efficacy and tolerability of palonosetron and granisetron in a Chinese population receiving highly emetogenic cisplatin-based chemotherapy or moderately emetogenic chemotherapy. Patients were stratified by chemotherapy with cisplatin (yes/no) and then randomly assigned to receive either palonosetron (0.25 mg i.v.) in the first cycle followed by granisetron (3 mg i.v.) in the second cycle or vice versa. The primary efficacy endpoint was the proportion of patients with complete response 0-24 h post-chemotherapy administration. The proportions of patients with complete response 24-120 and 0-120 h following chemotherapy were also compared. Of the 144 patients randomized, 36 (25%) received 60-80 mg/m(2) cisplatin; 66 of 72 patients in the palonosetron to granisetron group and 56 of 72 patients in the granisetron to palonosetron group completed treatment with both antiemetics. The efficacy and safety analyses included 128 palonosetron treatments and 138 granisetron treatments. Palonosetron consistently produced numerically higher complete response rates than granisetron in the acute phase (0-24 h, 71.09 vs. 65.22%), the delayed phase (24-120 h, 60.16 vs. 55.80%), and overall (0-120 h, 53.13 vs. 50.00%) though the differences were not significant. Both palonosetron and granisetron were well tolerated. Palonosetron was well tolerated and effective in preventing acute and delayed chemotherapy-induced nausea and vomiting in a Chinese population. When used as monotherapy, 0.25-mg palonosetron was not inferior to 3-mg granisetron for preventing vomiting following highly or moderately emetogenic chemotherapy.