Major advances in understanding the pathophysiology of acute leukemia have resulted in a dramatic increase in the availability of novel compounds for clinical trials. However, since the number of new drugs far exceeds the number of clinical trials that can be conducted because of the availability of eligible patients, there is an urgent need to utilize reliable preclinical models for the prioritization of the most promising potential therapies for those clinical trials. The most widely used preclinical models for the acute leukemias are human tumor xenografts established in immune-deficient mice, and genetically engineered mouse strains. This review summarizes the recent developments and considerations in the use of xenograft models of acute lymphoblastic leukemia, acute myeloid leukemia, and acute promyelocytic leukemia for the preclinical testing of new therapies.