The transplantable promyelocytic leukemia in BN rat (BNML) is a relatively slow-growing leukemia which has many characteristics in common with human acute myeloid leukemia (AML). As compared with normal myeloid cells, the leukemic cells have a prolonged transit time in the peripheral blood, a low growth fraction and high cell loss in the spleen and bone marrow, and an increasing doubling time towards the end stage of the disease. The normal hemopoiesis becomes progressively suppressed due to the disappearance of normal hemopoietic stem cells from the bone marrow. At the same time the number of normal stem cells in the peripheral blood and in the spleen increases to very high values, but this extramedullary hemopoiesis cannot adequately compensate for the decrease in the bone marrow. The exact proportion of clonogenic cells among the leukemic cells could not be determined due to a discrepancy between the median lethal tumor dose and the proportion of leukemia cells that produce colonies in the spleen. The survival time of rats receiving one or a few leukimogenic units is considerably longer than would be expected by extrapolation of the exponential growth that occurs between leukemic cell loads of 103 and 107. This initial lag period may be comparable to the prolonged remissions seen in a small proportion of treated patients, but the translation of leukemic cell loads and survival times from rat to man is still subject to a number of uncertainties. The BNML responds better to chemotherapeutic combinations that are effective in human AML than to regimes designed for human acute lymphoid leukemia. For the time being the BNML seems to offer the best opportunities for systematic studies of remission induction and drug selection which are relevant to the situation in human AML.