The present study examined the relationship between religious practice and psychological distress in a culturally diverse urban population to explore how religious affiliation, gender, ethnicity, and immigrant status affect this relationship. Data were drawn from a study of health care utilization in Montreal. A stratified community sample of 1485 yielded four religious groups: Protestant (n = 205), Catholic (813), Jewish (201), and Buddhist (150), and a group with no declared religion (116). The sample was composed of five ethnocultural groups: Anglophone Canadian-born, Francophone Canadian-born, Afro-Caribbean, Vietnamese, and Filipino immigrants. Psychological distress was assessed with the 12-item version of the General Health Questionnaire (GHQ). Religious involvement was measured with three items: 1) declared religion; 2) frequency of attendance at religious meetings; and 3) frequency of religious rituals performed at home. Multiple regression models examined the relationship of religious practice to distress, controlling for sociodemographic variables including ethnicity. Overall, attendance at religious services was associated with a lower GHQ score. Attendance at religious services also was inversely related to psychological distress for females, Protestants, Catholics, Filipinos, and Afro-Caribbeans; but not for males, Buddhists or Jews. Religious practice at home was not associated with level of distress for any group. The 'no declared religion' group had the highest mean GHQ score of all the groups. Results confirm the association between attendance at religious services and lower levels of distress, but reveal ethnospecific and gender effects indicating the need to understand the impact of religious practice on mental health in social and cultural context.