Purpose This study was designed to estimate the partial and complete response rates (CR and PR) of two novel drug pairs (vincristine and melphalan vs. ifosfamide and etoposide) and to improve overall survival of previously untreated patients with metastatic rhabdomyosarcoma. Patients and Methods One hundred twenty-eight patients were randomly assigned to phase II window therapy consisting of vincristine and melphalan (VM-containing regimen) or ifosfamide and etoposide (IE-containing regimen). Brief window therapy (12 wks) was immediately followed-up by vincristine, dactinomycin, and cyclophosphamide (VAC), chemotherapy, surgery, and irradiation, with continuation of either VM or IE in patients with initial response. Major endpoints were initial CR and PR rates after the phase II window phase of therapy, failure-free survival (FFS), and survival. Results Patients who received the VM-containing regimen experienced significantly more anemia, neutropenia, thrombocytopenia, and had more cyclophosphamide dose reductions. The initial PR and CR rates were not significantly different for patients treated with either regimen (VM, 74%; IE, 79%;P = 0.428). However, FFS and overall survival (OS) at 3 years were significantly better with the IE-containing regimen (FFS: 33% vs. 19%;P = 0.043; OS: 55% vs. 27%;P = 0.012). Conclusions Although the VM-containing regimen produced a high response rate, inclusion of melphalan appeared to limit the cyclophosphamide dose that could be administered, and ultimately, this regimen was associated with a significantly worse outcome than was the IE-containing regimen. Also, the IE-containing regimen was associated with a gratifyingly high survival rate at 3 years (55%), which is significantly higher than has been observed on any previous Intergroup Rhabdomyosarcoma Study Group regimen for similar patients. We believe that this promising outcome indicates that this drug pair merits further randomized testing in metastatic rhabdomyosarcoma.