A total of 120 patients were treated with granisetron either intramuscular (i.m.) or intravenous (i.v.) in a crossover design, over two successive cycles of moderately emetogenic chemotherapy. Of the 117 patients evaluable for efficacy, 74.4% receiving i.m. and 76.9% receiving i.v. treatment experienced a complete response (no vomiting, no more than mild nausea, no need for rescue medication and no study withdrawal in the 24 h following the onset of chemotherapy). Only a small proportion of the patients experienced any vomiting, either during the first 24 h or in the follow-up period of 4-10 days. There were no statistically significant differences in any of the efficacy parameters between the two routes administration of granisetron. Both formulations of granisetron were also equally well tolerated. The main treatment-related adverse effects were headache and constipation (experienced by 13-15% of patients); local reactions to i.m. injection of granisetron were experienced by 2.6% of patients.