One thousand eighty-six studies were reviewed to determine the prevalence, characteristics, and quality of measures of religion or religiosity in The Journal of Family Practice for the decade ending in 1986. Religious variables occurred at a relatively low rate, even among articles with some psychosocial content. The clear preponderance of religious measures pertained to patients rather than to providers, and religious variables were generally analyzed descriptively, not inferentially. In spite of an encouraging use of some measures of religious beliefs and practices, there remained a significant focus on denominational measures of religious status. These findings suggest that the consideration of religious variables in the family medicine literature has not been commensurate with the emerging picture of the role of religious commitment in health status and mental health status. It is suggested that increased emphasis on developing an original literature of record about religious variables in family medicine will promote the empirical assessment of the beneficial, neutral, and harmful effects of religion among family medicine patients and providers.