A growing body of research conducted by cognitive psychologists and behavioral economists finds that some features of human decision making are not adequately described by standard rational choice models of human behavior. In this paper, we offer experimental evidence suggesting that one of these features -- framing effects -- also arises in religious choices. When faced with a hypothetical choice between two religious options, respondents' choices were strongly influenced by the reference point from which they evaluated the options. Such results are difficult to reconcile with standard rational choice models, and they suggest alternatives to extant rational choice explanations of religious behavior. We discuss one such specific alternative: framing effects suggest that a "status quo bias" rather than "religious capital" may be responsible for relatively low levels of intergenerational mobility among religious denominations.