Background: Superiority of palonosetron to granisetron is uncertain, for patients with breast cancer receiving both steroid and NK1 inhibitor against CINV caused by AC regimen. Methods: We conducted a randomized double-blind active-controlled study. 341 chemo-naïve patients treated with AC regimen were randomized 1:1 to either (1) palonosetron 0.75 mg + dexamethasone + fosaprepitant or (2) granisetron 1mg + dexamethasone + fosaprepitant. Stratification factor was age ( < / > 55 yrs) and type of anthracycline (epirubicin / doxorubicin). Patients recorded episodes of emesis, nausea and rescue medication using a formed diary during 0 -120 hrs post-chemotherapy. Primary endpoint was complete response (CR = no emesis and no rescue medication) in delayed phase ( > 24 -120 hrs). Secondary endpoints include CR in acute (0-24 hrs) and overall (0 - 120hrs) phase. Nausea and emesis in acute, delayed, or overall phase was also evaluated. Treatment comparisons were performed using chi-square test with Yates’ continuity correction. Results: From Dec. 2012 to Oct. 2014, 326 evaluable patients were enrolled with comparable characteristics across groups; median age 54, 87 % of patients received epirubicin, 71 % of patients were light or no drinkers, 71 % of patients had no motion sickness, and 52 % of patients had morning sickness. CR rates were similar for patients treated with palonosetron and granisetron in delayed (62.3 vs 60.4 %; p = 0.8), acute (75.9 vs 73.2 %), and overall (54.9 vs 54.9 %) phase, respectively. Palonosetron reduced nausea (59.9 vs 72 %; p = 0.029) and emesis (10.5 vs 17.7 %; p = 0.088) in delayed phase. Adverse events occurred at similar rates across both groups. Conclusions: Although nausea and emesis in delayed phase was reduced, palonosetron at highest effective dose did not prove superiority to granisetron for prevention of CINV in patients with breast cancer receiving dexamethasone and fosaprepitant following AC regimen. Three drugs combination with steroid, 5-HT3 receptor antagonist, and NK1 receptor antagonist did not achieve CR in almost half of patients in this population. A new class of agents is needed. Clinical trial information: UMIN000008897.