In this paper, we discuss the cytokinetic basis for optimization of cancer therapy in humans. Specifically, we define a quantitative procedure for determination of the therapeutic acceptability of therapy schedule, discuss studies of the therapeutic gain that might result from treatment of acute lymphoid leukemia (ALL) with cytosine arabinoside (Ara-C), and indicate the kinds of cytokinetic information necessary for therapy scheduling. These studies suggest 1) that substantial therapeutic improvements may result from optimal therapy scheduling, 2) that therapy schedules selected at random are not likely to be beneficial, and 3) that consideration of the cytokinetic properties of the dose limiting normal tissues is critical during therapy design. We also discuss experimental studies of the response to Ara-C of the murine KHT sarcoma. These studies illustrate the substantial cytokinetic changes occur during therapy (e.g., a substantial increase in cell cycle traverse rate and almost complete recruitment) and show the importance of obtaining mid-treatment cytokinetic information about normal and malignant tissues.