In mid-July 2002, the Haute-Savoie Departmental Health and Social Affairs Directorate (Ddass) received a report of acute flu-like symptoms occurring in the Chamonix Valley. A descriptive epidemiological investigation was immediately launched by the Ddass and the Regional Health Agency (Cire) in collaboration with the French Institute for Public Health Surveillance (InVS). On August 19, the diagnosis of Q fever was confirmed, allowing for the immediate implementation of measures to inform and screen at-risk populations. The DDSV and the CNR for Rickettsia were notified, and epidemiological investigations were quickly conducted to identify the mode of transmission and the source of the outbreak. A probable case of Q fever was defined as a person residing in or visiting the Chamonix Valley or the municipalities located downstream of the valley during the month preceding the onset of symptoms and who, since June 1, 2002, had presented with a high fever and at least two of the following symptoms: headache, muscle pain, nausea, chills, and elevated transaminases. A confirmed case of Q fever was defined as a person with a positive Q fever serology (IgM phase II = 25) under the same geographic and temporal conditions as a probable case. A confirmed case presenting with flu-like symptoms was classified as a "clinical confirmed case." For the case-control study, controls were recruited by random selection from the telephone subscriber list and, in cases of absence from home or refusal at the time of the survey, by replacing them with a neighbor according to a random procedure. An indirect immunofluorescence test was performed for each control to exclude immunized individuals. The analysis included 26 cases and 102 controls. Over the entire period, 1,104 people were tested by the CNR. In the Chamonix Valley, 99 confirmed cases—including 78 clinical cases—and 33 probable cases were identified between June 14 and November 3, 2002. Among the confirmed cases, 12 were pregnant women and 3 were individuals with valvular heart disease. Sixteen people were hospitalized. No deaths were reported. Downstream in the valley, only 5 confirmed cases were identified. Two-thirds of the cases occurred during a 5-week period from early July to early August. The outbreak affected all municipalities in the valley. Incidence rates by residential neighborhood, calculated based on all confirmed and probable cases, identified three areas with the highest incidence located within the municipality of Chamonix. Analysis of exposure factors ruled out foodborne transmission. No differences were found between cases and controls regarding occupational and leisure activities or usual travel destinations. A statistically significant association was identified between the disease and having had close contact with or having attended a sheep transhumance. These human investigations confirmed the occurrence of a major Q fever epidemic confined to the Chamonix valley during the summer of 2002. The number of identified cases was likely underestimated due to the prevalence of asymptomatic cases and the difficulty in identifying cases among tourists—who were very numerous at that time—because of the long incubation period. The results of the descriptive and analytical investigations were consistent with the hypothesis of airborne transmission from livestock herds staying in or moving toward mountain pastures in the municipality of Chamonix. This mode of transmission makes it more difficult to identify the source, especially if it is mobile, as several sheep herds had stayed in the immediate vicinity of downtown Chamonix or had crossed the municipality via various routes. Following the human epidemiological investigations, veterinary investigations were conducted on farms that were permanently located in or had stayed in the Chamonix valley in order to take the necessary measures to prevent the emergence of new human cases. The investigation protocol included a descriptive epidemiological survey of the farms and, in a second phase, a serological and bacteriological survey aimed at identifying infected flocks and those shedding the bacteria. The survey identified 56 flocks belonging to 45 farmers. Since the descriptive epidemiological survey did not yield conclusive evidence to specifically suspect certain farms, all farms that had continued operations were subjected to the serological survey. In total, 17 farms (19 herds) tested positive, with 8 herds across 6 farmers exhibiting a seroprevalence exceeding 10%. The seroprevalence of farms located in the four municipalities of the Chamonix Valley was higher than that of farms located in the downstream municipalities, and it appeared that this significant difference originated in the sheep flocks. The bacteriological investigation focused on flocks with a seroprevalence greater than 10% or a non-zero seroprevalence that had stayed or moved through the epidemic zone, totaling 11 flocks (5 sheep, 4 cattle, and 2 goats). Vaginal and milk samples were collected from these flocks. Two sheep flocks were found to be shedding the bacteria in early 2003. Based on the summary of veterinary investigations, a score was established for each farm using various criteria for suspicion (date of lambing, locations and dates of manure spreading, locations where the flocks had been, and positive results from serological and shedding surveys). The two excreting sheep flocks were those with the highest scores. Although the bacterial excretion observed several months after the presumed period of human infections does not allow for the conclusion that these flocks were the source of the human outbreak, specific management measures were taken regarding these two farms with the aim of preventing a new outbreak. Comprehensive preventive measures were also implemented regarding the seropositive flocks. The occurrence of this type of human outbreak caused by a zoonosis underscores the importance of collaboration between human and animal health services in conducting investigations and implementing appropriate measures. It also led to the development of a veterinary investigation protocol to improve responsiveness in similar events and, in the absence of regulations, prompted a referral to AFSSA regarding risk assessment and measures applicable to products and animals. (R.A.)