We conduct a randomized evaluation of the effect of village-based schools on children's academic performance using a sample of 31 villages and 1,490 children in rural northwestern Afghanistan. The program significantly increases enrollment and test scores among all children, eliminates the 21 percentage point gender disparity in enrollment, and dramatically reduces the disparity in test scores. The intervention increases formal school enrollment by 42 percentage points among all children and increases test scores by 0.51 standard deviations (1.2 standard deviations for children that enroll in school). While all students benefit, the effects accrue disproportionately to girls. Evidence suggests that the village-based schools provide a comparable education to traditional schools. Estimating the effects of distance on academic outcomes, children prove very sensitive: enrollment and test scores fall by 16 percentage points and 0.19 standard deviations per mile. Distance affects girls more than boys--girls' enrollment falls by 6 percentage points more per mile (19 percentage points total per mile) and their test scores fall by an additional 0.09 standard deviations (0.24 standard deviations total per mile).