Ewing's sarcomas are characterized by recurrent chromosomal translocations expressing EWS-ETS fusion proteins, the most common of which is EWS-FLI.(1-5) EWS-FLI is an oncogenic transcription factor that regulates genes involved in tumorigenesis.(6,7) Because the Ewing's sarcoma cell of origin remains unknown, a variety of model systems have been developed to study EWS-FLI fusions,(8-14) and multiple microarray experiments describing potential EWS-FLI target genes have been reported.(8,10,11,13,15-21) Each model has potential benefits and drawbacks, but a large-scale comparison of these has not been reported. Herein we report a meta-analysis of the genes that are dysregulated by EWS-FLI in Ewing's sarcoma model systems. In general, EWS-FLI gain- and loss-of-function models in human cell types were well correlated to patient-derived tumor samples, while murine models were not. Using frequency analysis of dysregulated genes across multiple model systems, we identified a conserved "core" EWS-FLI transcriptional signature. This signature contained many of the genes known to be involved in the tumorigenic phenotype of Ewing's sarcoma, and also contained genes that have not been previously reported. Comparisons between the core EWS-FLI signature and published mesenchymal stem cell data support the recent assertion that mesenchymal stem cells are likely the Ewing's sarcoma precursor cell.(15) These results demonstrate the utility of using comparative analysis to validate model systems and emphasize the unique potential of this approach to identify both oncogenic and background cell signatures.