In the 2010s, Peru experienced an increase in public educational investment, a substantial improvement in public school learning outcomes, and an erosion in the private sector learning premium. We use longitudinal, geo-coded register data on primary schools and pupils in urban areas to study how the improvement in public schooling affected private schools. With a difference in differences (DiD) framework, we demonstrate that the increase in public school quality reduced enrolment and test scores in private schools, primarily in areas with lower education levels. A staggered DiD analysis shows that new public school openings also reduced enrolment in nearby private schools.