Household food insecurity is associated with multiple adverse outcomes in children and adolescents, including poor school performance. U.S. federal food assistance programs such as the Food Stamp Program (FSP) aim to help prevent household food insecurity and its outcomes. Program participation may act as a resource either to counteract the effects of constraints such as food insecurity or to modify the effects of food insecurity on outcomes. This study aimed to determine whether FSP participation was associated with child reading and mathematics learning, weight gain, and social skills and whether these associations depended on degree of estimated need for the program. Data used were from the Early Childhood Longitudinal Study-Kindergarten cohort, a large nationally representative sample. A difference (i.e., fixed-effects) model was used to minimize bias in the estimation of association resulting from unmeasured covariates. Starting FSP participation during the 4 years from K to third grade was associated with about a 3-point greater improvement in reading and mathematics score as compared with stopping FSP participation during that period. But it was for female students only that this association was large and significant. Children in households starting FSP participation had slightly but not significantly less weight gain compared with children in households stopping FSP participation. This study provides the strongest evidence to date that FSP participation plausibly has beneficial effects for children on nonnutritional outcomes, specifically academic learning. The mechanisms for this relationship are not well understood and may be through both dietary intake and stress.