Most studies on the effects of tracks during secondary education focus on specific short-term academic performance indicators. This study compared the effects of tracks on long-term unemployment from 1995 to 2015, using a Flemish longitudinal cohort study (4,333 students in 54 schools). Tracks in Flanders have a hierarchy in mean academic ability. Therefore, we compared pairs of hierarchically consecutive tracks by propensity score matching students across these tracks. Because many students changed track over time, we distinguished between students who followed a track completely and those who changed to a lower track. Discrete-time event history analysis was used to describe the probabilities of becoming unemployed when active (employed or in education) and the probabilities of becoming active when unemployed. The results showed that students who followed a higher track completely had lower probabilities of becoming unemployed.