While the benefits of additional schooling in the developing world are widely recognised, the best use of scarce resources to improve academic achievement remains unclear. We compare public investments in school infrastructure, school improvement grants, teacher qualifications, and attendance incentives on independently-gathered measures of academic skills as well as grade progression for 811year olds in india. We match a rich household survey containing a skills-assessment module, the india human development survey (ihds), with detailed measures of each districts education resources from the district information survey on education (dise). We also include border-pair fixed effects to control for unobserved heterogeneity. We find that incentives for children to attend school were associated with arithmetic, reading and writing skills, and grade progression. Investment in teachers were associated with greater probability a child could write and do more advanced math. Small improvement grants to schools were associated with better reading skills and writing ability. Investments in school infrastructure were only associated with improved writing ability. 2018, 2018 Informa uk limited, trading as taylor & francis group.