Women undergoing general anesthesia for breast surgery are at particular risk of postoperative nausea and vomiting. In a randomized, double-blinded, placebo-controlled trial, 90 patients scheduled for breast surgery, aged 33-63 years, received intravenously placebo, 3 mg granisetron, or 0.3 mg ramosetron (n = 30 of each) at the end of surgical procedure. A standard general anesthetic technique and postoperative analgesia were used. Emetic episodes and safety assessment were performed during 0-24 hours and 24-48 hours after anesthesia. The rate of patients experiencing emetic symptoms (nausea, retching, vomiting) 0-24 hours after anesthesia was 17% with granisetron (P = 0.013) and 10% with ramosetron (P = 0.002) compared with placebo (47%); the corresponding rate 24-48 hours after anesthesia was 27% (P = 0.032) and 7% (P = 0.001), compared with placebo (53%). In the 24-48 hours after anesthesia, there were fewer emetic episodes in patients who had received ramosetron than in those who had received granisetron (P = 0.039). The severity of nausea was less in patients receiving ramosetron than in those receiving granisetron (P = 0.044). Zero to 24 hours after anesthesia, no difference in the rate of patients having emetic symptoms and the severity of nausea was observed between the granisetron and ramosetron groups. The most common reported adverse events were headache and dizziness, and there were no difference in the incidence of adverse events due to the study drug among the 3 groups. In conclusion, prophylactic therapy with ramosetron is more effective than that with granisetron for the long-term prevention of postoperative nausea and vomiting in women undergoing general anesthesia for breast surgery.