Background/Context: Efforts to improve student achievement should increase graduation rates. However, work investigating the effects of student-level accountability has consistently demonstrated that increases in the standards for high school graduation are correlated with increases in dropout rates. The most favored explanation for this finding is that high-stakes testing policies that mandate grade repetition and high school exit exams may be the tipping point for students who are already struggling academically. These extra demands may, in fact, push students out of school. Purpose/Objective/Focus: This article examines two hypotheses regarding the relation between school-level accountability and dropout rates. The first posits that improvements in school performance lead to improved success for everyone. If school-level accountability systems improve a school for all students, then the proportion of students performing at grade level increases, and the dropout rate decreases. The second hypothesis posits that schools facing pressure to improve their overall accountability score may pursue this increase at the cost of other student outcomes, including dropout rate. Research Design: Our approach focuses on the dynamic relation between school-level academic achievement and dropout rates over time--that is, between one year's achievement and the subsequent year's dropout rate, and vice versa. This article employs longitudinal data of records on all students in North Carolina public schools over an 8-year period. Analyses employ fixed-effects models clustering schools and districts within years and controls each year for school size, percentage of students who were free/reduced-price lunch eligible, percentage of students who are ethnic minorities, and locale. Findings/Results: This study finds partial evidence that improvements in school-level academic performance will lead to improvements (i.e., decreases) in school-level dropout rates. Schools with improved performance saw decreased dropout rates following these successes. However, we find more evidence of a negative side of the quest for improved academic performance. When dropout rates increase, the performance composites in subsequent years increase. Conclusions/recommendations: Accountability systems need to remove any indirect benefit a school may receive from increasing its dropout rate. Schools should be held accountable for those who drop out of school. Given the personal and social costs of dropping out, accountability systems need to place more emphasis on dropout prevention. Such an emphasis could encompass increasing the dropout age and having the school's performance composite include scores of zero on end-of-grade tests for those who leave school.