This article aims to measure and compare the effectiveness of voucher schools and public schools on mathematics and reading test score trajectories from Grade 4 to Grade 10 in Chile. We analysed a national database that includes measures taken in 2007, 2011, and 2013 from a same sample of over 140,000 students. Multilevel-growth models were used to analyse the trajectories, and linear adjustment with propensity score matching was used to control for confounding variables. We found positive effects of voucher education on intercepts which, after controlling for schools’ socioeconomic composition and selection policies, became irrelevant in terms of effect sizes. Regarding slopes, voucher schools have a significant positive effect, which is irrelevant in terms of effect sizes if 1 year is considered, but relevant for a 12-year school trajectory. The results illustrate a stratification between the public and voucher systems during a period when intense neoliberal policies were enforced.