OBJECTIVE: To determine the prevalence of positive tuberculin tests in a population of patients requiring long-term care in a hospital setting. DESIGN: Prevalence study: to evaluate reaction to the test among patients who agreed to be included in the study. SETTING: The study took place in two units of the Enfant-Jésus hospital where patients are admitted for chronic care. PARTICIPANTS: A total of 108 patients were eligible for the study; 56 agreed to take part. One patient died before the study was completed. INTERVENTIONS: The tuberculin test consisted of an injection of PPD-T and reading the reaction 48 to 78 hours later. A reaction > 10 mm was considered significant. Patients with insignificant reactions were injected again 2 weeks later in order to evaluate positive response secondary to reactivation of the immune system (booster effect). MAIN OUTCOME MEASURES: Indications of previous tuberculosis, risk factors for tuberculosis, immunosuppressive medication, length of stay in hospital, size of reaction. RESULTS: Seventeen tuberculin tests were positive (30.9%); of these, six were positive after the second injection. CONCLUSION: The prevalence of positive tuberculin tests was high in our elderly population; this finding is comparable to the findings of American studies.