Study Objective: To compare the effectiveness of granisetron with placebo in the treatment of established postoperative nausea and vomiting (PONV). Design: Randomized, placebo-controlled study. Setting: 34 hospitals in Europe, Scandinavia, and South Africa. Patients: 519 ASA physical status I, II, and III patients who developed PONV within 4 hours of the end of surgery performed with general anesthesia. Interventions: Patients received a single intravenous dose of granisetron 0.1 mg, 1 mg, or 3 mg, or placebo when symptoms of nausea or vomiting were experienced. Additional rescue medication could be given at the investigator's discretion if nausea and vomiting were not controlled. Measurements and Main Results: At all doses investigated, granisetron was significantly more effective (p less than or equal to 0.001) than placebo in controlling vomiting: 38%, 46%, and 49% of patients receiving granisetron, 0.1 mg, 1.0 mg, and 3.0 mg, respectively, experienced no vomiting in the first 24 hours following drug administration, compared with 20% receiving placebo. There was a statistically significant linear relationship between vomiting confirmed the statistically significant difference between patients receiving granisetron and those receiving placebo. Granisetron was well tolerated: the most common adverse experiences were pain, constipation, anemia, and headache, and the incidence of adverse experiences was not statistically significantly higher in any of the granisetron groups than in the placebo group. Conclusion: Granisetron was significantly more effective than placebo in all groups. Further studies in specific subgroups may be warranted. (C) 1997 by Elsevier Science Inc.