This paper presents new evidence on the impact of bank branch expansion and credit access on human capital outcomes for children. Using a regression discontinuity design, we study a branch authorization policy by the Reserve Bank of India that encouraged banks to open branches in underbanked districts, where the population-to-branch ratio exceeded the national average. Bank presence, bank lending, and household borrowing increased. We find significant improvements in test scores: children in underbanked districts scored 0.16–0.22 SD higher on reading and math. We document three mechanisms. First, we find evidence for a demand-side channel where parents spent more on their children’s education and children spent more time on homework. Second, we document supply-side impacts in improvements in the quantity and quality of schools and teachers. Third, we find support for a labor market channel, with shifts away from agricultural employment and towards employment in manufacturing, while self-employed individuals expanded their businesses.