Soft tissue sarcomas are mesenchymal tumors which respond poorly to systemic therapy. Recent studies suggest a higher response rate with an increased doxorubicin dosage. However, this was parallel with a profound hematotoxicity in 75% of patients. Transfer of the human multidrug resistance 1 (MDR1) gene to normal hematopoietic stem cells and transplantation may significantly reduce the hematotoxicity of anthracyclin-based chemotherapy. To test this concept of supportive gene therapy in advance of a clinical study, we transduced mobilized peripheral blood progenitor cells (PBPC) with the retroviral vector SF91m3 containing the human MDR1 gene, transplanted these cells to immune-deficient mice, allowed 6 weeks for engraftment to occur and treated the animals with MDR1-based chemotherapy. In the MDR1-transduced group the human leukocytes were significantly protected from the toxicity of chemotherapy (p < 0.05). While the gene transfer rate was in the range of 10% and thus comparable to recent clinical trials, the gene expression was 59% of transduced cells and thus significantly higher than previously reported for less-advanced vectors. On the other hand, ifosfamide, a drug which has been used successfully for stem cell mobilization, is active in soft tissue sarcoma. Due to these favorable characteristics sarcoma is an attractive target to test the efficacy of MDR1 gene therapy in a clinical setting. Gene therapeutic strategies may also be used to directly target sarcoma cells, e.g. by transfer of suicide genes. We found that adenoassociated virus 2 (AAV-2) vectors efficiently transduce human HS-1 and HT1080 sarcoma cells (>90%) while other tumor cell lines and primary human PBPC were less susceptible. The thymidine kinase (TK) suicide gene was cloned into an AAV-2 vector and a complete kill of TK-transduced HS-1 and HT1080 cells was observed following exposure to aciclovir or ganciclovir (GCV), while >90% of mock-transduced HS-1 cells survived at these dosages. Transplantation of those sarcoma cells to nonobese diabetic (NOD)/LtSz-severe-combined immunodeficient (scid)/scid (NOD/SCID) mice resulted in a survival of >5 months in the AAV-TK-transduced/GCV-treated group, while the mice in the mock-transduced/GCV-treated group had died after 3 weeks. These data show that soft tissue sarcomas are a particularly suitable model system for the development and clinical testing of new gene therapeutic concepts.