This study was performed in a group of 89 patients treated for breast carcinoma with a FEC regimen (5-fluorouracil, epirubicin, cyclophosphamide). The aim was to evaluate the antiemetic efficacy of one 8 mg tablet ondansetron (OND), and in case of failure, to measure the recovering level by the administration of 8 mg i.v. OND. One tablet of OND was given before the first course. A complete or major control of emetic episodes (EE) (0-2 EE) over the 24 first hours was obtained in 55 patients (62%). A complete or major control of nausea (absent or mild) was obtained in 46 patients (52%). A success, defined as a complete or major control for EE and nausea during the first 5 days, was obtained in 36 patients (40%). These 36 patients were treated again with 8 mg oral OND at the second course, with a success level of 61% (22/36). Among the 53 patients unresponsive to the oral OND, 38 were treated by 8 mg i.v. OND at the second course. A complete or major control for EE and for nausea over the first 24 hours was obtained respectively in 11 patients (29%), and 11 patients (29%). No success was obtained. The 15 remaining patients, who had a very bad tolerance at the first course were excluded from the study and treated by OND-methylprednisolon, with only one success obtained. This study shows that the exclusive use of one 8 tablet OND is not sufficient in prevention of emesis induced by a FEC regimen, and that failures are only partially recovered by the i.v. route.