Following various indications suggesting an increase in leptospirosis cases in mainland France, particularly in the Dordogne, enhanced surveillance for this disease was implemented by the InVS in collaboration with the relevant Cire and Ddass offices, and an investigation was conducted into cases recorded by the CNR from August 18 to 29, 2003, as well as cases reported directly to the Ddass from June 1 to August 31. The identified cases were interviewed using a standardized questionnaire covering the clinical aspects of the disease and exposure to known risk factors for leptospirosis during the three weeks preceding the onset of the disease. This investigation was supplemented by a description of cases identified by the CNR from early to late August 2003 and a review of cases identified by the CNR in metropolitan France from January 1 to August 31, 2003, with a comparison to previous years. Of the 43 cases recorded, 39 (28 confirmed, 6 probable, and 5 suspected) were investigated. These cases had presented the first signs of leptospirosis between early June and late August, with 50% of them occurring between August 11 and 24. Thirty-seven cases were hospitalized, and 2 people died (1 confirmed case and 1 suspected case). Among the 28 confirmed cases, 5 serogroups were identified, predominantly (62%) Icterrohaemorrhagiae. Clusters of cases were identified in three departments: Aube, Ardennes, and Dordogne. Common or neighboring swimming sites were identified in Aube and Ardennes, and control measures (swimming bans) were implemented. In Dordogne, the reported cases were scattered throughout the department; two rivers were cited as swimming sites by two people, but at different locations. No common exposure was identified for the other cases in this department. The analysis of cases diagnosed by the CNR from June to August 2003 did not reveal any clusters of cases in time or space other than those in the three departments already mentioned. Furthermore, the serogroups were highly diverse. Finally, the findings regarding risk factors were consistent with the results of the study on leptospirosis risk factors in metropolitan France. A comparison of the number of cases observed in 2003 with those of the previous twenty years does not support the notion of a resurgence of leptospirosis during the summer of 2003. Following these investigations, it became necessary to develop a method for detecting alerts based on CNR data and a protocol for investigating (epidemiological and environmental) clusters of cases. Enhanced surveillance to better understand incidence and detect clusters of cases could be implemented in high-incidence regions, such as Aquitaine. It would also make it possible to assess the relevance and feasibility of adding leptospirosis to the list of notifiable diseases as a complement to the surveillance conducted by the CNR. Furthermore, public information on leptospirosis and individual prevention measures could be considered. (publisher’s abstract)