In the late 1980s, a series of federal laws were enacted which expanded Medicaid eligibility to more of the nation's children. States had a great amount of discretion in how fast and how far these expansions were implemented. As a result, there was great variation among the states in defining who was eligible for the program. This variation provides a rare opportunity to disentangle the effect of Medicaid from a child's socioeconomic status. Using data from the National Health Interview Survey, we address whether the Medicaid expansions improved the health and functional status of children. Econometric models were developed using fixed-effects regressions, and were estimated separately for white, black, and Hispanic children. White children experienced statistically significant reductions in acute health conditions and functional limitations. Black and Hispanic children showed some evidence of improved health conditions and functional status, but this evidence is inconclusive in the study sample. This may be due to differences in their access to appropriate health services or to the smaller sample size of minorities in each geographic area. The findings are also relevant to the implementation of the Children' Health Insurance Program (CHIP), the latest federal effort to expand access to health care to poor and near poor children. In many states, CHIP is being implemented in whole or in part through further Medicaid expansions. © 2002 by the Association for Policy Analysis and Management.