We investigated the antiemetic efficacy and safety of intravenous ondansetron infusion in the BMT setting. We conducted prospective randomized comparison trials between ondansetron at 2 dose levels and metoclopramide (MCP) plus droperidol for the prevention of chemotherapy-induced nausea and vomiting in 2 patient populations scheduled to undergo BMT. One patient population (n = 30) received CY alone, the other population (n = 30) received combination chemotherapy of Bu and CY. The CY alone group received ondansetron for 3 days, and the Bu/CY group received ondansetron for 7 days. The primary endpoints were emesis control and nausea. Secondary endpoints included acute (headache, diarrhea and sedation) and delayed (engraftment and regimen-related) side-effects. In both trials, ondansetron provided better emesis control than did MCP plus droperidol during CY administration (P = 0.009, 3-day trial; P = 0.0022, 7-day trial). There was a wide interpatient variation in serum ondansetron levels, although group averages were proportional to the dose administered. Intrapatient day-to-day variation was 10-30% and did not change significantly with concurrent CY administration. Antiemetic efficacy did not correlate with ondansetron serum levels at the doses tested. Headache incidence was similar in all groups. Sedation was highest in the MCP plus droperidol group (P = 0.048, 3-day trial; P = 0.016, 7-day trial). No statistically significant differences in engraftment or regimen-related toxicities were observed between groups in either trial. Ondansetron appears to be a safe and efficacious antiemetic during conditioning for BMT.