In this project, we construct a unique data set for Philadelphia, which merges information from vital statistics records with Medicaid claims records and individual-level data from publicly funded health centers. We then merge the individual-level data with a rich set of contextual information from multiple sources to create various measures of neighborhood quality, including housing and environmental quality, economic distress, social disorganization, health and service availability. We examine the influence of individual and contextual factors on racial and neighborhood variations on birth outcomes using rigorous statistical techniques suited for analyses of multilevel data. We will examine social class differences within race as well as the racial disparity itself. A second set of analyses addresses the effectiveness of publicly funded health services on prenatal care use, infant health, and costs for low-income women. We will compare the content of prenatal care and birth outcomes of women who received Medicaid services in a managed care setting versus women who received services from fee-for-service provides who accepted Medicaid patients in 1989-1992. Another set of analyses will focus specifically on women using city-funded health centers for prenatal care in 1993-1996. To complement the above analyses we will undertake an in-depth longitudinal study of low-income women. We propose to follow approximately 1,900 women from their first prenatal care visit through their baby s first birthday. This study will enable us to examine the effects of factors present in our conceptual framework that are not available from vital statistics data, including the mother s social support networks, living arrangements, housing stability, psychosocial factors, the wantedness of her pregnancy and risk-taking behaviors. The sample will consist of white, African American, Hispanic and Asian women. The sample composition will permit us to address explicitly the role of race in infant health, and to examine whether factors contributing to infant health are similar for low income women in all racial/ethnic groups.