Background: Brucellosis is a zoonotic disease subject to mandatory reporting in public health and to animal health control measures based on culling and vaccination. Efforts to control animal brucellosis enabled France to obtain the status of "officially free of bovine brucellosis" in 2005, and no cases of brucellosis in ruminants have been recorded in France since 2003. In this context, a survey of human cases of brucellosis diagnosed in France was conducted to identify residual risk factors and guide preventive measures in humans. Materials and methods: The descriptive survey focused on cases of brucellosis reported in France from June 2002 to June 2004. A case was defined as any patient presenting clinical signs compatible with the disease; for a confirmed case, this required bacterial isolation or a fourfold increase in anti-Brucella antibody titers in two samples taken at least 15 days apart by agglutination or indirect immunofluorescence, or seroconversion; and for a probable case, a single elevated serological titer and exclusion of differential diagnoses. For each case, clinical and laboratory data, as well as information on risk factors for brucellosis (consumption of raw milk products, contact with animals), were collected using a standardized questionnaire. Results: During the study, 105 reports were received, including 72 cases and 26 false cases (25%) that did not meet the case definition. The annual incidence of brucellosis was 0.05 per 100,000 inhabitants. Among the cases, 32% had a focal clinical presentation, most commonly arthritis. The diagnosis was established by bacterial isolation in 49 cases (65%). 80% of the cases were imported (infected during travel to an endemic area or through consumption of cheese brought back from an endemic area). The main countries of infection for the cases were Portugal (n=14), Algeria (n=7), and Turkey (n=6). Cases had significantly more contact with animals in endemic areas than false cases (p=0.05) and had more frequently consumed imported dairy products than false cases (p=0.004). Conclusion: Our study confirmed the effectiveness of the animal health control measures implemented. With the eradication of animal brucellosis in France, the incidence of human disease has sharply declined, and the majority of cases are now imported. In this context of very low prevalence and numerous false cases, direct diagnosis should be prioritized to limit the number of false cases. To improve the quality of surveillance, recommendations are made to increase the specificity of surveillance, in particular to consider only cases diagnosed by bacterial isolation or a rise in serological titer. (R.A.)