Malignant transformation is likely to arise in a subset of organ-specific primitive cells that are subverted to acquire the properties of uncontrolled self-renewal.1,2 It is therefore likely that stem cells and tumor-initiating cells share many properties and that an understanding of the biology of normal stem cells and the identification of the pathways and molecules that regulate their self-renewal may result in our ability to design inhibitors that control the growth of tumor cells.In an interesting recent paper, Lukacs et al.3 show that the polycomb group transcriptional repressor, Bmi-1, regulates the self-renewal of normal prostate stem cells and also contributes to the initiation of prostate cancer.