Introduction. Between January 2 and 4, 2006, six cases of Legionnaires’ disease were reported to the Departmental Directorate of Health and Social Affairs (DDASS) of Ille-et-Vilaine. The cases were observed to be clustered in time and within a geographic area limited to the city of Rennes, suggesting a common source of infection. Methods. A descriptive epidemiological investigation was conducted to validate the presumption of an outbreak and guide environmental investigations in search of common sources of Legionella exposure. A case was defined as any person who, since December 11, 2005, pneumonia with laboratory confirmation of Legionnaires’ disease caused by Legionella pneumophila serogroup 1 (LP1) and who resided, worked, or had visited the Rennes metropolitan area (the city of Rennes and neighboring municipalities) within 10 days prior to the onset of clinical symptoms. Environmental investigations primarily focused on a geographic area common to the cases, identified based on their movements and their homes. The investigations focused on wet cooling towers (WCTs), known to be the installations most at risk of widespread dissemination of aerosols contaminated with Legionella. Any unregistered WCTs were sought out. The operating conditions and self-monitoring results for these installations were examined. In addition to these investigations, individual exposures to Legionella were investigated. The domestic hot water systems in the cases’ homes were described. Similarly, the operating conditions of the WRTs were assessed in an adjacent geographic area frequented by only one of the cases. The clinical and environmental strains isolated during the investigations were compared by the National Legionella Reference Center using pulsed-field gel electrophoresis after DNA macro-restriction. Results. Eight cases were included in this outbreak: the dates of onset of the first symptoms ranged from December 21, 2005, to January 1, 2006. The period from December 19 to 27 appears to be a particularly high-risk period during which the majority of cases may have been infected. Two cases died. Review of self-monitoring results and additional analyses identified three TARs where Legionella risk control had failed. Self-monitoring at one of the TARs (facility 1) identified an environmental strain identical to the single clinical strain isolated. This was the "Rennes" strain identified during a previous epidemic episode in Rennes in 2000. Environmental risk management. The inspection services for classified facilities monitored the implementation of environmental risk management measures at each of the three TARs. Discussion. The shape of the epidemic curve suggests the existence of a common source of exposure. This hypothesis is supported by the fact that no other individual exposure to Legionella was identified. The results of the environmental and microbiological investigations suggest a link between the contamination of a water treatment plant located east of Rennes and the occurrence of some of the cases. The contamination of the facility by the "Rennes" strain on December 21, 2005, is the main argument in favor of such a link. However, the evidence gathered during the investigations remains insufficient to confirm that the 8 clustered cases were attributable to this single facility. This new outbreak suggests that the "Rennes" strain may be endemic, as it had previously been identified only during the 2000 outbreak. Furthermore, the low number of cases included in this outbreak may reflect improvements in surveillance and collaborative efforts. Indeed, the investigations conducted immediately upon the reporting of the first cases made it possible to monitor and supplement control measures and likely to bring this outbreak under control quickly. Conclusion and recommendations. This episode underscores the importance of more systematically combining positive Legionella diagnoses made by antigen testing with bacteriological samples. It confirms the importance of early collaboration among all partners involved in Legionella risk management. (R.A.)