Primary human AML cells can be isolated and studied in vitro, but many experimental questions can only be addressed using in vivo models. In particular, tractable animal models are needed to test novel therapies. The genetic complexity of human AML poses significant challenges for the generation of reliable animal models. The hematopoietic systems of both zebrafish ( Danio rerio) and Drosophila have been well characterized ( reviewed in [5, 31]) . Both organisms are well suited to forward genetics mutagenesis screens. Although this approach has been useful for identification of mutants with hematopoietic phenotypes ( e.g., cloche), the impact on cancer biology and hematopoietic malignancies in particular has been limited. A zebrafish model of acute lymphoblastic leukemia has been generated [37] and Drosophila models have shed light on the biology of epithelial tumors ( reviewed in [60]). Nonetheless, in vivo modeling of human AML relies most heavily on mice. Most cellular, molecular, and developmental features of the hematopoietic system are well conserved across mammalian species. The availability of the human and mouse genome sequences and the capability of manipulating the mouse genome make mice the most valuable model organism for AML research. Mice have additional practical value because they have a short reproductive cycle and are relatively inexpensive to house.