An administrative rule in North Carolina allowed students who failed an exam to retake it approximately two weeks later, triggering a brief yet intense test preparation period. We develop a structural model that accounts for selection and find that these students score much higher on the retest. Using a regression discontinuity design, we find substantial fade-out of the test score gains after one year but some persistence thereafter. Unlike other interventions that produce similar initial increases in performance, we do not observe benefits to long-term outcomes. Our findings highlight that persistence should be accounted for when comparing educational interventions.