### Abstract Past research has shown that intergroup contact can be a promising intervention to improve intergroup relations and that contact-based interventions might be most effective during adolescence. In post-conflict Northern Ireland, widespread residential segregation and a largely separate school system limit opportunities for intergroup contact between adolescents from the Catholic and Protestant communities. We evaluated whether a large-scale intervention to facilitate intergroup contact between students attending separate schools (the ‘Shared Education’ program) improves a range of outcomes relevant for intergroup relations in Northern Ireland. We conducted a five-wave longitudinal, quasi-experimental study that followed a large sample of school students (*N* = 5,159, *M* = 12.4, age range: 10–14 years; 2,988 girls, 2,044 boys) from 56 predominantly Catholic or Protestant schools from sixth to tenth grade. We compared the developmental trajectories of students who, in ninth (14–15 years) and tenth (15–16 years) grade, shared some classes with students from the other community, as part of the program, to students who did not. We found that participating in shared classes had a medium-size, positive effect on the amount of intergroup contact students had outside of class, and small, positive effects on students’ outgroup attitudes, outgroup trust, and intergroup empathy (but not on their intergroup anxiety, future contact intentions, deprovincialization, or multicultural beliefs). Our findings show that a school-based program of shared education can provide a viable and effective intervention to facilitate intergroup contact, improve intergroup relations, and foster social integration among adolescents at a large scale in a post-conflict society.