PROJECT SUMMARY African American and Latino populations have worse adult health than their White peers (hereafter “health disparities”). These health gaps often begin in childhood, and have persisted, in part, because of continuing inequalities in the social determinants of health. Far less research has identified population-level interventions that level the playing field on the social determinants of health across population subgroups. The scant evidence on effective population-level interventions is a significant barrier to reducing health gaps. Rigorous prior studies have revealed the plausibility for large-scale nutrition policies to reduce racial/ethnic gaps in children’s academic performance—a strong predictor of adult educational attainment, which is one of the most fundamental determinants of health. These studies have illuminated the intermediary mechanisms in the pathway between school nutrition policies, (junk) food availability in communities near schools and children’s academic performance. No longitudinal studies have thoroughly investigated the population-level influences on academic performance of nutrition policies together with modifiable characteristics of nearby-school neighborhoods. This longitudinal study capitalizes on a series of natural experiments generated by California’s policies to improve nutrition standards for foods and drinks sold to children in schools, and the changes in nutrition standards for school meals put in place by the Healthy, Hunger Free Kids Act of 2010. The study will determine (a) the effectiveness of population-level school nutrition policy interventions in improving children’s academic performance and reducing related gaps among African American and Latino vs. White peers; (b) the extent to which modifiable, food-related factors in communities near schools influence longitudinal changes in academic performance disparities; and (c) if school nutrition policy effectiveness varies by those community factors. To generate valid inferences about the population-level effectiveness of these policies on academic performance, this study uses the strongest possible approach to evaluate non-randomized exposures: a difference-in-differences analysis that includes within-child change in exposures. Difference in difference analyses will also examine the effects of child-level changes in modifiable characteristics of communities near their schools and changes in academic performance (and disparities). This study is unparalleled because we use powerful longitudinal data on academic performance among a diverse population of 11.8 million children. Policy and community interventions in and around schools hold potential to reduce disparities given their community-oriented focus, existing infrastructures and networks that facilitate their large-scale implementation and broad reach. Given its robust design, the study will have a significant impact on evidence-based nutrition policy and population-level interventions to...