A large literature shows that religious participation is associated with a wide range of behaviors and outcomes, but what drives this association is unclear. On the one hand, this association may stem from correlations in preferences, where those with tastes for religion coincidentally have particular tastes for other behaviors as well. Alternately, religious participation may directly affect behavior; for example many religious organizations impose rules and proscriptions on their members and these rules may affect members' decisions. Using the canonical economic model of religiosity, I develop an empirical test to investigate the importance of religious proscriptions on behavior. Several empirical applications of this test are conducted; the results indicate a strong role for religious proscriptions in determining behavior. The test developed here does not require an instrumental variable for religion and could be applied to the study of criminal gangs, terrorist organizations, fraternities, communes, political groups, and other "social clubs."