Sarcomas are a heterogeneous group of mesenchymal malignancies and unfortunately there are limited functional genomics platforms to assess the molecular pathways contributing to sarcomagenesis. Thus, novel model systems are needed to validate which genes should be targeted for therapeutic intervention. We hypothesized that delivery of oncogenes into mouse skeletal muscle using a retroviral (RCAS-TVA) system would result in sarcomagenesis. We also sought to determine if the cell type transformed (mesenchymal progenitors vs. terminally differentiated tissues) would influence sarcoma biology. Cells transduced with RCAS vectors directing the expression of oncoproteins KrasG12D, c-Myc and/or Igf2 were injected into the hindlimbs of mice that expressed the retroviral TVA receptor in neural/mesenchymal progenitors, skeletal/cardiac muscle or ubiquitously (N-tva, AKE and BKE strains respectively). Disrupting the G1 checkpoint CDKN2 (p16/p19-/-) resulted in sarcoma in 30% of p16/p19-/- xN-tva mice with a median latency of 23 weeks (range 8-40 weeks). A similar incidence occurred in p16/p19-/- xBKE mice (32%), however, a shorter median latency (10.4 weeks) was observed. p16/p19-/- xAKE mice also developed sarcomas (24% incidence; median 9 weeks) yet 31% of mice also developed lung sarcomas. Gene-anchored PCR demonstrated retroviral DNA integration in 86% of N-tva, 93% of BKE and 88% of AKE tumors. KrasG12D was the most frequent oncogene isolated. Oncogene delivery by the RCAS-TVA system can generate sarcomas in mice with a defective cell cycle checkpoint. Sarcoma biology differed between the different RCAS models we created, likely due to the cell population being transformed. This genetically flexible system will be a valuable tool for sarcoma research.