This paper studies donor and nonprofit responses to the star rating system designed by Charity Navigator. I find that charities respond to the rating system by changing their behavior to move above the star thresholds, leading to “bunching.” This response is equivalent to charities reducing spending on administration by half, partially driven by charity misreporting. Moreover, donors reward charities based on these ratings, such that crossing the threshold from 3- to the highest 4-star rating raises contributions by 6%. These results highlight the importance of optimal rating design, including charity incentives, cognitive shortcuts, and the need for monitoring.