This paper uses a quasi-natural policy experiment in Germany, the G8 reform, to examine the impact of schooling intensity on student learning. The G8 reform compresses secondary school for academic-track students from nine to eight years, while holding fixed the overall academic content and total instruction time required for graduation, resulting in a higher schooling intensity per grade. Using German extension of the Programme for International Student Assessment data, we find that this reform improves test scores on average, but the effect differs across subgroups of students. The reform effect is larger for girls than for boys, for students with German-born parents than for those with immigrant parents, and for students having more books at home. The heterogeneous reform effects cannot be explained by changes in observed channels. Instead, quantile regression results suggest that unobserved heterogeneity plays an important role: Whereas high-performing students significantly improve their test scores, the lowest-performing students hardly improve at all after the reform. We interpret the unobserved heterogeneity as reflecting students' capability to cope with the increase in schooling intensity.