Objective: This study examined how well recipients of psychiatric or substance abuse services self-report their care. Methods: Self-reported use of behavioral health services, as reported by 1,012 respondents to a survey mailed to Minnesota Medicaid managed care enrollees receiving services with a behavioral health diagnosis other than severe mental illness in 2000, was compared with Medicaid agency records. Results: Overall, 15 percent of respondents said they had not received services, although administrative data indicated otherwise. In bivariate analyses, failure to report treatment was associated with gender, age, education, and diagnosis but not with race or ethnicity or with residence. The effects of gender and diagnosis were insignificant after analyses controlled for age and education. Conclusions: Self-reports of behavioral health service use are a relatively accurate method of obtaining information on use. The finding that underreporting of receipt of behavioral health services was not random suggests that reporting errors could introduce bias when comparing groups that differ on age or education.