OBJECTIVE: To describe a systematic evaluation of the outcomes associated with revising institutional guidelines for the prevention of acute chemotherapy-induced nausea and vomiting (CINV) to promote cost-effective use of the serotonin (5-HT₃) antagonists. METHODS: The 5-HT₃ antagonist of choice in the antiemetic guidelines was revised from intravenous ondansetron to oral granisetron in August 1995. Patient assessments were conducted immediately prior to (Period 1) and after (Period 2) guideline revision using validated questionnaires. The effectiveness of the two 5-HT₃ antagonists were compared and reported to the prescribing oncologists. Outcomes were assessed one year after guideline revision (Period 3) using identical methods. RESULTS: No difference was found in the rate of total control (no emesis, no nausea) between patients receiving oral granisetron (60%) and intravenous ondansetron (56%) (p = 0.408, Period 1 vs. 2). Nausea severity, the number of emesis episodes, and use of rescue antiemetics were also equivalent. Prescriber compliance with using the 5-HT₃ antagonist of choice and dose increased from 48% to 61% following adoption of oral granisetron. By Period 3, compliance increased to 78%, and satisfactory control of acute CINV was again documented. The costs for prevention of acute CINV decreased from $107 in Period 1 (intravenous ondansetron only) to $65 in Period 3 (oral granisetron). CONCLUSIONS: Outcomes associated with use of oral granisetron and intravenous ondansetron were equivalent in this patient population. Guideline revision and outcome documentation by the oncology pharmacists resulted in increased compliance with institution guidelines and a 40% cost savings.