This study aimed to determine the incidence of tuberculosis among hospital employees in four major urban institutions, which employ nearly 90% of all hospital workers caring for adult patients in a city of 2.5 million inhabitants. It also had the objective of finding out whether this incidence changes according to the in-hospital setting, i.e. the departments, and thus, whether tuberculosis can be considered to pose an occupational risk. The study population consisted of all healthcare workers of the four hospitals between 1986 and 1998. From these, those who developed tuberculosis within this period were determined and were considered as the case group. All the remaining employees constituted the control group. Fifty-nine healthcare workers were found to have developed the disease. The annual incidence was found to range between 0.016 and 0.139%. Tuberculosis was seen 3 times more frequently among the hospital employees than the general population. The workers in departments of chest diseases were found to have a higher risk than those of other departments (OR: 6.37, CI: 3.69-11.00). Similarly, the nurses were also at a higher risk than the doctors (OR: 2.63, CI: 1.12-6.36). These findings suggest that tuberculosis can be considered as an occupational disease.