The antiemetic efficacy of ondansetron and dexamethasone (Ondex) was randomly compared to that of high-dose metoclopramide, dexamethasone, and orphenadrine (Control) in the prevention of emesis induced by cyclophosphamide-doxorubicin chemotherapy in 64 chemotherapy-naive breast cancer patients. For the control of acute emesis (day 1), patients were randomized to receive either ondasetron 8 mg p.o. 1 hour prior to chemotherapy (CT) and repeated after 6 and 12 hours plus dexamethasone 20 mg i.v. 40 minutes prior to CT (Ondex) or dexamethasone 20 mg i.v. 40 minutes prior to CT, orphenadrine 40 mg i.m. 35 minutes prior to CT and metoclopramide 3 mg/kg i.v. 30 minutes prior to CT and repeated after 90 minutes followed by 40 mg p.o. every 3 hours for 4 times (Control). To control delayed emesis, patients on Ondex received ondansetron 8 mg PO t.i.d. days 2 and 3 and patients in the Control arm received metoclopramide 0.5 mg/kg p.o. q.i.d. and dexamethasone 8 mg i.m. b.i.d. days 2 and 3. Complete and major control of acute emesis was observed in 74%/94% and 44%/67% of patients treated with Ondex and Control, respectively (p < .01/p < .005). Acute nausea was absent in 38% and 34% of patients treated with Ondex and Control, respectively (p = NS). Complete and major control of delayed emesis (days 2-5) was observed in 65%/91% versus 44%/66% of patients in the Ondex and Control arms, respectively (p = NS/p < .01). In patients receiving 6 courses of FEC/FAC, control of acute emesis was significantly superior with Ondex at all treatment courses.