When might religious belief lower ethical standards? We propose a theory of religion and immorality that makes 3 central predictions. First, people will judge immoral acts as more permissible when they make divine attributions for these acts, seeing them as enabled by an intervening God. Second, people will be more likely to make divine attributions when evaluating passive immorality (e.g., keeping a lost wallet) than active immorality (e.g., pick-pocketing) because human action makes people less likely to infer God's agency. Third, believers will be more likely than nonbelievers to perpetrate passive immorality, because they feel justified taking advantage of God's beneficence. Thirteen studies support these predictions. Our findings show that people who attribute events to God judge morally questionable behaviors more leniently (Study 1), American states with more prayer groups have higher rates of crime (Study 2), and experimentally manipulated divine attributions lead people to see selfish and harmful behavior as less immoral (Study 3). Divine attributions-and corresponding moral permissibility-are more likely with passive immorality than with active immorality (Studies 4-7). Compared with nonbelievers, believers are more likely to justify their own passive immorality (Study 8), and to commit everyday acts of passive immorality such as parking across multiple spaces (Study 9) and keeping overdue library books (Study 10). A novel behavioral economics task reveals that although passive immorality is not affected by religious priming, it does correlate with self-reported religious belief (Studies 11-13). Finally, an internal meta-analysis supports our predictions. (PsycINFO Database Record (c) 2019 APA, all rights reserved).