What are the effects of sociodemographic factors on the use of outpatient mental health services when different demographic groups have identical health insurance coverage? The authors answer this question using data from the Rand Health Insurance Experiment. Health insurance was randomly assigned to families representative of the nonaged, noninstitutionalized civilian population in six U.S. sites. Income has no significant total effect on use when all income groups have the same coverage. When the effects of variables correlated with socioeconomic status are removed, users with higher socioeconomic status are significantly more likely to choose a mental health specialist rather than only general medical providers for their mental health care (P less than 0.05); among those who visit mental health specialists, those with higher socioeconomic status incur significantly greater expenses (P less than 0.10). Women use significantly more mental health services than men (P less than 0.05), who in turn use significantly more mental health services than children (P less than 0.05), even after controlling for demographic factors, health status, and insurance coverage. Similarly, there are large differences (roughly sixfold) by site in outpatient mental health expenses even when all sites have identical coverage.