The empirical evidence concerning the relationship of psychiatric disorders to health care utilization and costs is mixed and primarily retrospective. Therefore, a case-control study was conducted to prospectively examine the association of psychiatric disorders with health care utilization in an adult primary care internal medicine patient population, controlling for the effects of medical morbidity, adverse events, age, race, gender, employment status, and health insurance coverage. Samples of primary care high utilizer (HU; 125 men, 125 women with primary care visits in 1998 above the 95th percentile) vs. mid-range utilizer (MU; 125 men, 125 women; two primary care visits in 1998) patients were compared using archival automated medical record data from the index year (1998), and from the following year (1999) for prospective analyses. HU (compared to MU) participants were younger, had higher medical and psychiatric morbidity, and had higher levels of outpatient specialty medical care utilization. In multivariate analyses, (a) anxiety disorder diagnoses uniquely contributed to identifying HU patients, and (b) after controlling for initial primary care utilization status, anxiety, depressive and addictive disorders were prospectively associated with medical illness complexity and primary, specialty, and emergency medical care utilization. Although behavioral health disorders (including addictive as well as depressive and anxiety diagnoses) and high utilization of primary health care services are related, these prospective findings suggest that behavioral health disorders make an independent contribution to non-psychiatric health care utilization.