As the major publicly financed early childhood education and care program, Head Start provides comprehensive and high-quality services, including early education, medical, dental, mental health, nutrition, parental involvement, and other social services, for disadvantaged preschoolers in the U.S. It aims to improve participants' school readiness, social competence, learning skills, health, and nutrition. Due to the issue of selection bias, the effects of Head Start have been a subject of continuous debate.

This dissertation investigates the effects of Head Start participation in the year prior to kindergarten on child outcomes from kindergarten to elementary school years, and examines whether the effects, if any, vary by participants' race/ethnicity, English speaking at home, or gender. It adopts multiple analytic approaches, including propensity score matching, fixed effects, and multivariate regressions, and uses data from the Early Childhood Longitudinal Study, Kindergarten Class of 1998–99 (ECLS-K). This dissertation includes three free-standing empirical papers focusing on different aspects of participants' outcomes respectively: test scores, childhood overweight and obesity, and social-emotional development. Implications for future research, practice, and policy changes are discussed.

Overall, the dissertation finds that compared to non-parental non-center-based care (i.e., exclusively parental care, relative care, and non-relative care), Head Start tends to lift children's test scores in math and reading, reduce their risk of overweight/obesity through the 3rd and 5th grade years, and increase children's externalizing problem behaviors in kindergarten but not the 3^rd grade or later years. Compared to other center-based care Head Start is less likely to increase children's test scores in kindergarten but not the 3^rd grade or later years, more likely to reduce the risk of childhood overweight, obesity, and problem behaviors, and more likely to increase children's self-control ability and interpersonal skills. Compared to multiple non-parental care arrangements, Head Start tends to increase children's test scores, approaches to learning, self-control, and interpersonal skills and to reduce their risk of overweight/obesity, and externalizing problem behaviors. Overall, the dissertation finds little evidence that the effects of Head Start vary significantly by child race/ethnicity, English speaking at home, or gender.