We study the effects of electrification on educational outcomes in Peru by taking advantage of a program that rapidly increased electricity coverage in rural areas. Using household survey panel data from 2007-2010, we document that: i) girls living in treated districts are more likely to be enrolled in school, and this effect is larger among younger girls; ii) this positive effect on female enrollment does not translate into higher attendance rates; iii) households in treated areas spend more money on younger girls' education. In addition, we use school-level panel data from 2007-2012 on Math and Reading test scores to show that treatment is associated with a reduction in learning. But, among treated schools, longer treatment exposure increases scores in Reading for boys and girls; and improves performance in Math, only among boys. Finally, our estimates are robust to controlling for other confounding interventions.