Purpose We evaluated the noninferiority of dexamethasone (DEX) on day 1, with sparing on days 2 and 3, combined with neurokinin-1 receptor antagonist (NK1-RA) and palonosetron (Palo) compared with the 3-day use of DEX in highly-emetogenic chemotherapy (HEC). Patients and Methods Patients who were scheduled to receive HEC (cisplatin ≥ 50 mg/m2 or anthracycline plus cyclophosphamide) were randomly assigned to receive either DEX on days 1 to 3 (Arm D3) or DEX on day 1 and placebo on days 2 and 3 (Arm D1) combined with NK1-RA and Palo. The primary end point was complete response (CR), defined as no emesis and no rescue medications during the overall (0 to 120 h) phase. The noninferiority margin was set at -15.0% (Arm D1 - Arm D3). Results A total of 396 patients-196 and 200 patients in Arms D3 and D1, respectively-were evaluated. CR rates during the overall period were 46.9% for Arm D3 and 44.0% for Arm D1 (95% CI, -12.6% to 6.8%; P = .007). CR rates during the acute (0 to 24 h) phase were 63.3% and 64.5% for Arms D3 and D1, respectively (95% CI, -8.1% to 10.6%; P < .001), and they were 56.6% and 51.5%, respectively, during the delayed (24 to 120 h) phase (95% CI, -14.8% to 4.6%; P = .023). Hot flushes and tremors were observed more frequently as DEX-related adverse events on days 4 and 5 in Arm D3, whereas anorexia, depression, and fatigue were observed more frequently on days 2 and 3 in Arm D1. As an indication of quality of life, global health status was similar in both arms. Conclusion Antiemetic DEX administration on days 2 and 3 can be spared when combined with NK1-RA and Palo in HEC.