Programs linking college aid to academic achievement could work either by lowering the cost of college or by inducing additional student effort. I examine the PROMISE program in West Virginia, which offers free tuition to students who maintain a minimum GPA and course load. Using administrative data, I exploit discontinuities in the eligibility formula and the timing of implementation to estimate causal effects. I find robust and significant impacts on key academic outcomes. Impacts are concentrated around the annual requirements for scholarship renewal, suggesting that the program works via incentives for academic achievement, not simply by relaxing financial constraints. (Contains 40 footnotes, 6 tables and 8 figures.)