Background: The significance of repeated treatment with the 5-HT3 receptor antagonist for prophylaxis of chemotherapy-induced emesis remains to be clarified. Patients and Methods: A retrospective analysis was performed to compare the effects of single and repeated treatment with granisetron on anorexia, nausea and vomiting in patients with breast cancer who undertook anthracycline and cyclophosphamide-based cancer chemotherapy. Results: The control of anorexia was significantly better in the single treatment group than in the repeated treatment group (54% versus 73%; odds ratio (OR), 0.433; 95% confidence intervals (CI), 0.226-0.828; p=0.016), although the rate of complete response to any signs of the gastrointestinal side-effects was not different between the two groups (37% versus 39%; OR, 0.911; CI, 0.489-1.700; p=0.874). However, the incidence of constipation was more frequent in the repeated treatment group (60% versus 37%; OR, 2-586; CI, 1.388-4.818; p=0.003). Conclusion: Repeated treatment with 5-HT3 receptor antagonist is not likely to be beneficial to breast cancer patients who undertook anthracycline/cyclophosphamide combination chemotherapy.