In a large randomized controlled trial, we test the hypothesis that incentives for physical activity can improve academic performance. We found strong support for this hypothesis: University students who were incentivized to go to the gym had a significant improvement in academic performance, by, on average, 0.15 standard deviations compared to a control group that did not receive any incentives. The success of this indirect incentive for academic performance emphasizes the importance of non-cognitive skills in achieving academic goals. Students who were incentivized to exercise report improved self-control and a healthier life-style.  Overall, the study demonstrates that incentivizing exercise can be an important tool in improving educational achievements.