Clinical trials with laboratory correlates were conducted in patients with acute leukemia to determine if relationships exist between drug dose and growth of surviving leukemia cells. The therapeutic design was based on findings in the leukemic rat that relate the initial dose of drug and tumor kill to the magnitude of residual tumor proliferation and sensitivity to a second drug. Patients with acute myelocytic leukemia received cytarabine, either 2 or 6 g/m2/72 h by continuous infusion. The presence and magnitude of change between initial and residual tumor after treatment, as measured by change in labeling indices, depended on the "priming" dose of drug. The amount of perturbation correlated with clinical response to cytarabine given at the time of induced proliferation. With results which parallel the rat data, the direct relationship of initial drug dose and proliferation of residual tumor is demonstrated in humans, and lends support to the design of our clinical trials of timed sequential therapy.