PURPOSE: We studied the frequency, causes, and predictors of adverse events in children with acute lymphoblastic leukemia (ALL) who had completed treatment on contemporary clinical protocols between 1984 and 1999. Our goal was to use the information to further refine therapy and advance cure rates. METHODS: Cumulative incidence functions of any post-treatment failure or any post-treatment relapse were estimated by the method of Kalbfleisch and Prentice and compared with Gray's test. The Cox proportional hazards model was used to identify independent prognostic factors. RESULTS: Of the 827 patients who completed all treatment while in initial complete remission, 134 patients subsequently had major adverse events, including 90 leukemic relapses, 40 second malignancies, and four deaths in remission. The cumulative incidence of any adverse event was 14.0% +/- 1.2% (SE) at 5 years and 16.9% +/- 1.4% at 10 years. The risk of any leukemic relapse was 10.0% +/- 1.1% at 5 years and 11.4% +/- 1.2% at 10 years. Male sex was the only independent predictor of relapse (hazard ratio, 1.74; 95% CI, 1.11 to 2.74; P = .02). CONCLUSION: Further treatment refinements for children with ALL should aim not only to decrease the leukemic relapse rate, but also to reduce the risk of development of second malignancies.