OBJECTIVE: Admissions to an acute care psychiatric hospital were studied to determine what proportion of admissions were for HIV-related psychiatric disturbances, what evidence exists that the HIV-related admissions constitute a new population of psychiatric inpatients, and whether any psychiatric variables distinguish patients with HIV-related disorders from patients without such disorders. METHODS: Charts for 2,094 consecutive admissions to the hospital in a three-year period were reviewed. Chi square analyses and t tests were used to compare various groups of HIV-related admissions and non-HIV-related admissions on sociodemographic characteristics, length of stay, history of psychiatric hospitalization, and other clinical measures. RESULTS: One- hundred and sixty-three admissions (7.8 percent of the total admissions) were judged to be HIV related. The largest category of HIV- related admissions, nearly half, were patients experiencing functional or psychological complications of HIV infection or risk. AIDS-phobic patients, who expressed irrational concerns about infection, accounted for 20.2 percent, and patients with organic manifestations for 18.4 percent. If the AIDS-phobic, HIV-bereaved, and factitious HIV-positive admissions were excluded, the rate of HIV-related admissions was 5.2 percent. CONCLUSIONS: Psychiatric hospitals can expect a significant but not overwhelming number of HIV-related admissions.