Genetically defined models of cancer represent one of the most important tools in cancer biology. These models are a critical resource for investigating the basic pathophysiology of cancer initiation and development, as well as for the identification of novel therapeutic approaches. However, a significant shortcoming of current models is that nearly all genetically defined models are created in non-human organisms, such as mice and zebrafish. Although the importance of these organisms cannot be understated, it is also apparent that there are many instances where human (cancer) biology diverges from mouse and zebrafish biology. Consequently, the lack of genetically defined, human models of cancer is a critical barrier to understanding basic cancer biology and developing new therapies. A significant difficulty in developing cancer models in human cells is the frequent inability to combine a specific genetic lesion (mutation) with the correct cell type at the appropriate developmental stage (cell-of-origin). Our laboratory is focused on developing innovative approaches to creating genetically defined, as well as biologically relevant, models of cancer in human cells by combining genomic data from the sequencing of cancer genomes with the pluripotent nature of human embryonic stem cells. Using this approach, we have generated a genetically defined model of Ewing sarcoma, a common pediatric tumor, from human stem cells differentiating through an embryoid body intermediate. The goal of our work is to now use these model cells to study the pathophysiology of Ewing sarcoma tumors, as well as develop additional human tumor models through more advanced stem cell differentiation methodologies.