The aim of this study was to test the hypothesis that brucellosis in Italy is a food-borne, rather than an occupational disease. This hypothesis was tested using data for both human and animal populations from the period 1997-2002. The correlation between the distribution of the disease in the human, sheep and goat populations was analysed, as were the risk factors for the disease, with respect to gender, age, occupation and residence of the individuals involved. Notifications of human brucellosis, which are mandatory in Italy, reach a peak between April and June. However, considering the standard incubation period of 2-4 weeks, and the fact that lamb slaughter is traditionally at a peak during the Easter period, it might be expected that occupational exposure would result in a peak of human cases between March and May. The observed peak between April and June could be related to the production and consumption of fresh cheese, starting just after lamb slaughter. The age of patients showed a fairly uniform distribution, and analysis of incidence rates of human brucellosis between 1997 and 2002 showed that the incidence rates were consistent with an occupational exposure risk of about 25%.