Between July 31 and August 5, 2003, 12 cases of Legionnaires' disease caused by Legionella pneumophila serogroup 1 were reported to the Hérault Departmental Directorate of Health and Social Affairs (DDASS), five of whom had visited Montpellier only once for a very brief period. An epidemiological investigation was initiated by the DDASS and the Southern Interregional Epidemiology Unit (CIRE), in collaboration with public health surveillance partners, with the objectives of describing the outbreak and identifying the common source(s) of infection. The initial results of this investigation tended to rule out a nosocomial source and directed the environmental investigation toward other sources of contamination, such as cooling towers (CTs). A case was defined as any person who, since July 15, 2003, had presented with pneumonia and had laboratory confirmation of Legionnaires’ disease caused by Legionella pneumophila serogroup 1 (isolation, seroconversion, positive urinary soluble antigen, or serological titer greater than or equal to 1/256) [1] and who was residing, working, or had visited Montpellier—even briefly—within 10 days prior to the onset of symptoms. A "single-visit" case was defined as a case not permanently residing or working in the city of Montpellier, having visited Montpellier only once within 10 days prior to the onset of symptoms. Active case finding was implemented among healthcare professionals in the Hérault department. To identify any cases of Legionnaires’ disease involving individuals who had stayed in Montpellier since July 15, 2003, other DDASS offices and the European Network for the Surveillance of Travel-Related Legionellosis (EWGLI) were alerted. A survey of Tars in the Montpellier metropolitan area was conducted by the Regional Directorate for Industry, Research, and the Environment (Drire), the Montpellier Municipal Health and Hygiene Service (SCHS), and the Ddass. Control samples were taken from these facilities. Other potential sources of contamination investigated included: the city’s water distribution network, domestic hot water at the patients’ homes, and decorative fountains in the city center. To assess the possible role of weather in the dispersal, survival, and proliferation of Legionella, several models comparing the epidemic curve with meteorological conditions were tested. The clinical and environmental Legionella strains isolated were sent to the National Reference Center (CNR) for Legionella in Lyon for molecular typing. A total of 31 cases of Legionnaires’ disease caused by Legionella pneumophila serogroup 1 were identified, including 8 “single-visit” cases. The first case appeared on July 15 and the last on August 12, 2003. All had resided, worked, or visited downtown Montpellier within the 10 days preceding the onset of illness. The 8 "single-visit" cases had visited Montpellier on July 21, 2003, and had frequented a limited area of downtown. Among the clinical samples from the patients, 8 strains were isolated and shared the same genomic profile (Legionella pneumophila serogroup 1). Samples taken after the alert in the Tar region within the area frequented by the cases revealed that 15 establishments were contaminated with Legionella. Five exceeded 10⁶ CFU/L, including two within the area frequented by all "single-visit" cases. Eleven different strains of Legionella pneumophila serogroup 1 were identified, including five in the same public fountain. However, none of the strains were identical to the human strain. Samples taken from the city’s water systems and decorative fountains were in compliance with regulations. No Legionella pneumophila serogroup 1 was found in the homes of the cases. The various models tested comparing the epidemic curve with meteorological conditions showed that there was a significant correlation between variations in outdoor air humidity levels and the occurrence of cases. (R.A.)