Biological factors and therapy interact in complex ways to affect prognosis of children with acute lymphocytic leukemia. Therefore, it is important that trials prospectively collect data on potential prognostic factors (such as age, white blood cell count, DNA index, cytogenetics, immunophenotype, central nervous system status, and early treatment response) in all patients. As results of treatment improve, subsequent trials must be large enough to detect small differences in outcomes. Results should be reported after sufficient follow-up, using multivariate analyses, and in a format that permits comparison with outcomes at other centers. Attention to the above will permit an approach to treatment that adapts the intensity of therapy to the risk of relapse.