This prospective randomized study compared the efficacy and toxicity of granisetron and dexamethasone to those of granisetron alone for antiemetic control in patients receiving high-dose chemotherapy with or without total body irradiation (TBI) for stem cell transplantation. Patients were divided randomly into 2 groups. Groups received granisetron twice daily at a dose of 40 microg/kg with or without 4 mg dexamethasone (GS group and G group, respectively), starting 30 minutes before each dose of chemotherapeutic agent or TBI, or 12 hours after the first dose if TBI or a drug was given once a day. Fifty patients were evaluated for the analysis. During the first 24 hours of conditioning, 23 of 25 patients (92.0%) in the GS group achieved complete control of emesis (no emetic episodes over the course of a day), compared with 72.0% in the G group (P = .06). For patients receiving TBI on the first day of conditioning, complete emetic control was achieved in all patients (100.0%) in the GS group, compared with 63.2% in the G group (P = .02). The same degree of emetic control was maintained throughout the conditioning period in 38.8% of the GS group and 29.9% of the G group (P = .10). Adverse reactions were observed more frequently in the GS group (68.0% versus 5.0% in the G group). These reactions included insomnia, headache, flushing, and hyperglycemia. None of the events were serious. We conclude that granisetron with dexamethasone seems superior to granisetron alone for the prevention of emesis resulting from the conditioning regimen; however, the more frequent side effects may limit the wide use of this combination.