Substantial previous research indicates that married persons enjoy better mental and physical health than others. One prominent explanation for these patterns is that married persons practice better health behaviors, including lower levels of alcohol consumption, particularly heavy drinking. However, studies of marital status differences in health behaviors have ignored the role of religious homo/heterogamy and attendance, important potential confounding and moderating factors. Focusing on alcohol behavior, we analyze 1977–1994 General Social Survey data, finding that lower levels of drinking are especially pronounced among homogamous conservative couples, and to a lesser extent among mixed-faith unions with one conservative partner. Controlling for other background characteristics, nonconservatives married to nonconservatives do not consistently consume less alcohol than unmarried persons, suggesting that much of the apparent marital status effect on drinking is actually due to religious factors. Contrary to some prior theory and research, these patterns are broadly consistent for both men and women.