We evaluate a 'girl-friendly' primary school program in Burkina Faso using a regression discontinuity design. After 2.5 years, the program increased enrollment by 19 percentage points and increased test scores by 0.41 ...