A B S T R A C T PurposeTo examine population-based improvements in survival and the impact of clinical covariates on outcome among children and adolescents with acute lymphoblastic leukemia (ALL) enrolled onto Children's Oncology Group (COG) clinical trials between 1990 and 2005. Patients and MethodsIn total, 21,626 persons age 0 to 22 years were enrolled onto COG ALL clinical trials from 1990 to 2005, representing 55.8% of ALL cases estimated to occur among US persons younger than age 20 years during this period. This period was divided into three eras (1990-1994, 1995-1999, and 2000-2005) that included similar patient numbers to examine changes in 5-and 10-year survival over time and the relationship of those changes in survival to clinical covariates, with additional analyses of cause of death. ResultsFive-year survival rates increased from 83.7% in 1990-1994 to 90.4% in 2000-2005 (P Ͻ .001). Survival improved significantly in all subgroups (except for infants age Յ 1 year), including males and females; those age 1 to 9 years, 10ϩ years, or 15ϩ years; in whites, blacks, and other races; in Hispanics, non-Hispanics, and patients of unknown ethnicity; in those with B-cell or T-cell immunophenotype; and in those with National Cancer Institute (NCI) standard-or high-risk clinical features. Survival rates for infants changed little, but death following relapse/disease progression decreased and death related to toxicity increased. ConclusionThis study documents ongoing survival improvements for children and adolescents with ALL. Thirty-six percent of deaths occurred among children with NCI standard-risk features emphasizing that efforts to further improve survival must be directed at both high-risk subsets and at those children predicted to have an excellent chance for cure.