Unlike many other high-stakes exams which test students only at the end of a course, the English A levels use a modular format where students are allowed to resit past units within the course. Since the introduction of resits for A levels in 2000, the A level results have been improving significantly and resits are seen as the main reason for the big increases. Using the findings of research based on the accounts of students and teaching staff, this article examines some of the less obvious effects of resits on student learning than that reflected by exam results in terms of students' approach to learning. It concludes that allowing unlimited resits in high-stakes exams could produce some undesirable consequences for student learning and could be unfair and unfit for the purposes of the exam. (Contains 4 tables, 5 figures, and 3 notes.)