Academic dismissal policies are increasingly implemented to promote academic success, with existing empirical evidence mostly restricted to short-run outcomes. This study examines long-term academic outcomes of academic dismissal for two cohorts (N=1707) of first-year bachelor students in Economics and Business at a Dutch university. Using administrative records, regression discontinuity design estimates suggest that academic dismissal does not relate to a difference in the propensity of graduation, nor to a change in study delay, when comparing students around the academic dismissal threshold. Not meeting this credit-threshold forces students to leave, and most decide to re-enroll in the same (43.4%) program elsewhere or at least within the academic domain (41.9%). Thus, while academic dismissal forces students to switch, its intended purpose of redirecting students to a different field of study is not observed. Implications for why academic dismissal might not deliver on the intended efficiency or effectiveness gains are discussed.