We use recently released, nationally representative data from the National Health Interview Survey-Multiple Cause of Death linked file to model the association of religious attendance and sociodemographic, health, and behavioral correlates with overall and cause-specific mortality. Religious attendance is associated with U.S. adult mortality in a graded fashion: People who never attend exhibit 1.87 times the risk of death in the follow-up period compared with people who attend more than once a week. This translates into a seven-year difference in life expectancy at age 20 between those who never attend and those who attend more than once a week. Health selectivity is responsible for a portion of the religious attendance effect: People who do not attend church or religious services are also more likely to be unhealthy and, consequently, to die. However, religious attendance also works through increased social ties and behavioral factors to decrease the risks of death. And although the magnitude of the association between religious attendance and mortality varies by cause of death, the direction of the association is consistent across causes.