Ifosfamide and doxorubicin are the most active agents in the treatment of sarcomas and are characterized by a marked dose-response relationship. The objective of this study was to determine the maximum tolerated dose (MTD) of both agents in combination under granulocyte-macrophage colony-stimulating factor (GM-CSF) cover. PATIENTS AND METHODS: Thirty-three patients with untreated sarcomas (soft tissue: n = 20; gynecological: n = 11; bone: n = 2) were treated with ifosfamide 12 g/m2 by continuous i.v. infusion over five days and doxorubicin with dose escalation from 50 mg/m2 i.v. bolus divided on two days, then to 60 mg/m2 bolus divided on three days. Ifosfamide was reduced to 10 g/m2 and doxorubicin was further escalated up to 90 mg/m2. GM-CSF (5 micrograms/kg/day subcutaneously) was started 24 hours after chemotherapy and continued for 10 days. RESULTS: The MTD was reached with the combination of ifosfamide at 12 g/m2 and doxorubicin at 60 mg/m2. But with ifosfamide 10 g/m2 and doxorubicin 90 mg/m2 the MTD was not obtained. While severe leukopenia and granulopenia were observed at all-dose levels, severe anemia was more frequently related to the highest dose of ifosfamide. Severe thrombopenia and mucositis were more commonly observed at the highest dose of doxorubicin. Ifosfamide 10 g/m2 and doxorubicin 90 mg/m2 induced WHO grade 4 leukopenia in 58%, grade 3-4 thrombopenia in 42%, and anemia in 31% of cycles. Mucositis was minor in 50% of cycles. The overall response rate among 31 evaluable patients was 55% (95 confidence interval (CI): 36%-73%), with four (13%) complete responders and 13 (42%) partial responders. Response rates based on soft-tissue sarcomas or gynecological sarcomas alone were similar. Ten patients could be treated by elective surgery and/or radiotherapy. The total group of patients reached a median survival of two years, with 25% (SE 8%) survivors after three years. CONCLUSIONS: The dose level of ifosfamide 10 g/m2 and doxorubicin 90 mg/m2 with supportive GM-CSF is manageable in a multicenter setting and should be further tested in regular phase II trials, including patients with gynecological and soft-tissue sarcomas. Transient toxicity with myelosuppression should be accepted in order to obtain a high antitumor activity of this regimen and a potential improvement in survival.