Recent advances in cytogenetics and molecular genetics have made it possible to identify an array of genomic abnormalities with prognostic and therapeutic significance. Hyperdiploidy > 50 chromosomes and ETV6-CBFA2 fusions have been used to identify low-risk cases, and BCR-ABL and MLL-AF4 to define high-risk leukemias. Despite their clinical utility, the risk classification system based on these findings lack absolute precision and should be complemented with other variables, the most important of which is the early blast cell response to remission induction therapy. Studies of tumor suppressor genes and proto-oncogenes in the BCL2 family genes may unravel the mechanisms of leukemia cell progression and the development of drug resistance, leading to innovative therapies. As the cure rates for childhood acute lymphoblastic leukemia (ALL) approach 80%, precise methods of risk assessment are needed to permit better selection of treatment that is neither excessive nor inadequate for individual patients. Because one or more genetic abnormalities underlie every case of leukemia, a risk assignment system based on primary genetic abnormalities has great intuitive appeal. Even though over 90% of childhood ALL cases can be readily classified according to numerical or gross structural chromosomal abnormalities, molecular analyses are essential to identify therapeutically relevant, submicroscopic genetic lesions not visible by karyotyping. This review focuses mainly on recent advances in genetic studies that have contributed to therapeutic advances or that hold promise for the future.