We know surprisingly little about the influence of race-blind school admissions on student outcomes. This paper studies a unique reform where a large, urban school district was federally mandated to adopt a race-blind lottery system to fill seats in its oversubscribed magnet schools. The district had previously integrated its schools by conducting separate admissions lotteries for black and non-black students to offset the predominantly black applicant pools. The switch to race-blind lotteries dramatically segregated subsequent magnet school cohorts. I show that race-blind admissions caused the more segregated schools to enroll students with lower average baseline achievement and to employ lower value-added teachers due to sorting. I also find that segregation is further exacerbated by "white flight" as white students transfer out of the district after attending more segregated schools. Ultimately, the mandated segregation decreases student standardized test scores and four-year college attendance. I provide suggestive evidence that the impact of racial segregation is partially mediated by changes to peer baseline achievement.