This study compared the effects of being in different tracks during the first 3 years of secondary education on student academic performance. A sample of a longitudinal cohort study in Flanders (3,205 students in 46 schools) was used to describe the learning gains for mathematics and reading comprehension. Four tracks were distinguished, with a clear hierarchy in mean student academic ability. A comparison was made per pair of tracks that are hierarchically consecutive by matching students who are comparable across these tracks. Three matching methods were applied: propensity score matching, Mahalanobis distance matching and coarsened exact matching. Multilevel latent growth curves and generalized estimating equations were used to describe learning gains over the first 3 years of secondary education. The results reveal that hierarchically higher tracks benefit students' academic performance for mathematics and reading comprehension. However, effect sizes differ across tracks, time, and outcomes. Moreover, differences in relative learning progress between compared tracks diminish over time.