Many patients admitted to psychiatric institutions have a history of risk factors for contamination with the hepatitis B and C viruses (HBV and HCV). Immunization policies for psychiatric institution patients are not widely know. A study conducted in a 600-bed psychiatric institution in Paris, France, to evaluate the proportion of potentially contaminating patients at admission, as well as immunization use in HBV-negative patients. Serologic markers for the HBV and HCV were looked for prospectively in all patients admitted over a two-month period. Immunization use was evaluated in the patients who had a first serologic test in 1995 or 1996 and follow-up tests until the end of 1997. The prospective part of the study demonstrated preadmission exposure to the HBV in 23.0 +/- 6.0% of patients. This proportion was larger in men (26.0%) than in women (14.8%), although the difference was not statistically significant (P = 0.10). Four patients (2.0%) tested positive for the HbS Ag. Among the HBV-negative patients, 13.0% received the vaccine; all had protective antibody levels. These patients were younger and more likely to be first-time admissions (18.4% vs 3.6%, P < 0.01). HCV seroprevalence was 6.0%. Serologic tests for the HBV were requested for 327 patients between January 1995 and december 1996. Among the patients who were seronegative at admission and received follow-up at the study hospital, only 13.8% were immunized at this hospital. Of the HCV-positive subjects, 63.3% were also HBV-positive. None of the HCV-positive HBV-negative subjects received immunization subsequently. Six to eight per cent of patients admitted to the study hospital are potentially contaminating for the HBV or HCV. The level of hepatitis B vaccine use is too low, particularly in high-risk patients. These data indicate a need for policies aimed at effectively preventing HBV and HCV transmission (information, education, immunization campaigns), both during the psychiatric institution stay and after discharge.