A rich experimental literature demonstrates positive effects of pay-per-visit fitness incentives. However, most insurance plans that provide fitness incentives follow a different structure, offering membership reimbursements conditional on meeting a specific attendance threshold. We provide the first evidence in the literature on gym incentives of this structure, exploiting the introduction and subsequent discontinuation of a large-scale wellness program at a major American university. Our analysis leverages individual-level administrative data on gym attendance for the universe of students over a five-year period: the three years that the policy was in place, one year before implementation, and one year after termination. This provides us with 100,000 student-year observations and 1.5 million gym visits. Using bunching methods and difference-in-difference designs, we provide four empirical results. First, we document that the policy led to significant bunching at the attendance threshold. Second, we show that the program increased average gym visits by almost five visits per semester, a 20% increase from the mean. Third, we find that the policy not only motivated students who were previously near the threshold, but rather increased attendance across the entire visit distribution. Finally, we show that approximately 50% of the effect persists a year after program termination. Taken together, these results suggest that rebate-framed incentives with a high attendance threshold can induce healthy behaviors in the short-term, and that these positive behaviors persist even after the incentives have been removed.