The cancer stem cell hypothesis postulates that tumor growth is driven by a rare subpopulation of tumor cells. Much of the supporting evidence for this intriguing idea is derived from xenotransplantation experiments in which human leukemia cells are grown in immunocompromised mice. We show that, when lymphomas and leukemias of mouse origin are transplanted into histocompatible mice, a very high frequency (at least 1 in 10) of the tumor cells can seed tumor growth. We suggest that the low frequency of tumor-sustaining cells observed in xenotransplantation studies may reflect the limited ability of human tumor cells to adapt to growth in a foreign (mouse) milieu.