In this paper I explore the psychology of ritual performance and present a simple graphical model that clarifies several issues in William Irons’s theory of religion as a “hard-to-fake” sign of commitment. Irons posits that religious behaviors or rituals serve as costly signals of an individual’s commitment to a religious group. Increased commitment among members of a religious group may facilitate intra-group cooperation, which is argued to be the primary adaptive benefit of religion. Here I propose a proximate explanation for (...) how individuals are able to pay the short-term costs of ritual performance to achieve the long-term fitness benefits offered by religious groups. The model addresses three significant problems raised by Irons’s theory. First, the model explains why potential free-riders do not join religious groups even when there are significant net benefits that members of religious groups can achieve. Second, the model clarifies how costly a ritual must be to achieve stability and prevent potential free-riders from joining the religious group. Third, the model suggests why religious groups may require adherents to perform private rituals that are not observed by others. Several hypotheses generated from the model are also discussed. (shrink)