OBJECTIVE To determine the prevalence of psychiatric disorders among patients with syncope; the relationship between psychiatric disorders and the cause of syncope; and the outcome of syncope patients with psychiatric disorders. METHODS In a prospective cohort study in a university-affiliated medical center, hospitalized and ambulatory patients with syncope underwent a standardized medical and psychiatric evaluation consisting of the Diagnostic Interview Schedule (DIS) sections for somatization, panic, generalized anxiety, major depression disorders, and drug and alcohol dependence and abuse disorders. RESULTS Of 414 patients who completed the DIS, 82 (20%) met the criteria for at least one of the psychiatric disorders or alcohol/drug disorders. Patients with at least one of the four major psychiatric disorders (n = 58) were more likely to report > or = 4 syncopal events in the past year (odds ratio (OR) 2.5, P < 0.04); and to have prodromal symptoms prior to syncopal events (OR 2.4, P < 0.04) than were patients without psychiatric disorders. Syncope patients with only alcohol/drug disorders (n = 24) were more likely to be male (OR 7.7, P < 0.001) and younger than 65 years old (OR 4.5, P < 0.001) than patients without alcohol/drug disorders. One-year recurrence rate for syncope in patients with any psychiatric diagnosis was 35%, compared to 15% in those without a psychiatric disorder (P < 0.0001). Physicians were able to recognize potential psychiatric or alcohol/drug disorders in 70% of patients with more than one psychiatric disorder, but failed to note or treat 60% of the patients with only one of the assessed disorders. CONCLUSIONS We recommend screening for psychiatric disorders in patients with unexplained syncope, especially in individuals with recurrent syncope and multiple physical symptoms, or for males who are under 65 years of age.