OBJECTIVE: This study investigated how insurance coverage for mental health services affects outpatient mental health service utilization among those with and among those without a DSM-III psychiatric diagnosis. The authors used a representative community sample to compare the regression effects of insurance coverage on utilization of mental health services among these subjects. METHOD: Data are from the second wave of the Piedmont, North Carolina, site of the Epidemiologic Catchment Area project. These data contain DSM-III diagnostic measures derived from the National Institute of Mental Health Diagnostic Interview Schedule as well as measures of insurance coverage and utilization. Responses from 2,889 community residents were analyzed using both ordinary least squares and logistic regression. RESULTS: In both models, insurance coverage was strongly associated with care among those with as well as among those without a psychiatric disorder. The association between coverage and the probability of care was strongest among those with a disorder. CONCLUSIONS: The findings are not consistent with the claim that failing to provide insurance coverage will reduce discretionary but not necessary mental health care utilization. They provide evidence that failing to provide insurance coverage will reduce utilization as much or more among those with a psychiatric disorder as among those without. This result has important implications for health care reform.