Objectives The effects of religiosity and sociodemographic characteristics on support for same‐sex marriage (SSM) are estimated. Methods An ordered probability model with ordinal endogenous treatment is estimated. Treatment effects of religiosity and marginal effects of other covariates are calculated. Results Religiosity reduces the likelihood of SSM support. This impact can be attenuated by educational achievement, Democratic Party affiliation, higher incomes, and greater contact with gays and lesbians. The importance of religion has no differential impact on the marginal effects of other covariates. Conclusions We expect the litigation of cases in which gay married couples allege victimization from some form of discrimination, which, in turn, will be defended on the grounds of religious freedom. Religious freedom supporters will come from the intrinsically religious, evangelicals, strong Republicans, and African Americans; same‐sex couples will draw support from those who are more educated, strongly Democratic, and at the higher end of the income distribution.