Acute lymphoblastic leukemia (ALL) is the most common cancer in children. Careful building upon past clinical trials and thoughtful application of our limited knowledge of pharmacology have provided steady improvement in outcome for newly diagnosed patients. Precise identification of the many patients who are unlikely to relapse with current effective regimens is required to avoid the morbidity of further intensification of therapy. Progress is sorely lacking for relapsed patients. Most patients who relapse die. Gene expression arrays and comparative genomic hybridization have further extended our appreciation of the known immunophenotypic and genetic diversity of childhood ALL. Insight into the molecular mechanisms of treatment failure may provide guidance for future efforts.