The capacity of filgrastim to reduce the myelotoxicity of a 16-week intensive chemotherapy regimen has been investigated in 24 operable breast cancer patients with > or = 10 metastatic axillary nodes. Five patients were treated with chemotherapy alone (control group); 19 patients were treated with chemotherapy and filgrastim, 5 microg/kg/day s.c. Six patients in the latter group were treated from day 4 to day 7 (level 1), seven from day 10 to day 13 (level 2), and six from day 4 to day 7 and day 10 to day 13 (level 3). A total of 135 courses were administered: neutropenia was the most severe toxicity, and the prophylactic use of filgrastim does not reduce its severity. Moreover, the dose intensities of antiblastic drugs actually received by the patients were not significantly different in the four study groups. Among the patients treated at level 3, there were three toxic deaths: one patient died because of febrile neutropenia and sepsis, two patients because of ischemic colitis. At a median follow-up of 15 months, 17 patients were alive, and 15 patients were disease free. The use of filgrastim does not ameliorate myelotoxicity and does not allow the administration of the planned doses of antiblastic drugs of a 16-week intensive chemotherapy regimen.