We studied the proliferative activity of leukemic cells obtained from the peripheral blood and bone marrow of 34 patients; 30 with acute leukemia and 4 with chronic myelogenous leukemia in blastic crisis. Flow cytometry was performed using monoclonal antibody against DNA polymerase alpha. Since fresh and frozen cells showed virtually identical DNA polymerase alpha-positive populations and flow cytometric histograms, 52 cryopreserved samples (25 from peripheral blood and 27 from bone marrow) were used in this study. The DNA polymerase alpha-positive population ranged from 20.4% to 84.7% in peripheral blood, and from 6.5% to 92.1% in bone marrow. A positive correlation (r = 0.76, P < 0.01) was found between DNA polymerase alpha-positive populations in peripheral blood and bone marrow from the same patient. This suggests that the DNA polymerase alpha-positive population in the bone marrow can be estimated from that in peripheral blood. No relationship was observed between the positive population and the response to chemotherapy. Statistical analyses for all cases showed no relationship between the DNA polymerase alpha-positive population and either the tumor cell count or time to reach a nadir. However, a negative correlation was observed between the positive population in bone marrow samples and the time to reach a nadir (r = -0.64, P < 0.05) in those patients who achieved a complete response. In addition, in the cases of acute non-lymphocytic leukemia who did not respond to chemotherapy, a positive correlation was observed between the tumor cell count in bone marrow and the DNA polymerase alpha-positive population (r = 0.93, P < 0.01). Thus, the method described here provides a simple and time-efficient means of detecting the proliferative activity of leukemic cells, which is a useful parameter in the treatment of leukemia.