Staff at the Paris Observatory became aware of what they considered to be a high number of cancer cases among employees in one of the buildings at the Meudon site between 1995 and 2001. The Occupational Health Department of the French Institute for Public Health Surveillance was asked in December 2001 to investigate this suspected excess of cases, in collaboration with the facility’s occupational health physicians. The reported cases were: - fourteen cases among women (total = approximately 43): one adenocarcinoma, one basal cell skin cancer, one cancer of the ear canal, two generalized cancers (primary site not identified), two melanomas, five breast cancers, two thyroid cancers. - eight cases among men (total = approximately 80): one basal cell carcinoma, one lymphoma, two multiple myelomas, one pancreatic cancer, two lung cancers, one prostate cancer. The initial incidence calculation* was based on 11 reported cases among women, 7 cases among men, and an approximate description of the employee population* (43 women and 80 men). It showed a potential excess of cancer cases among women, with a Standardized Incidence Ratio* (SIR = ratio of observed cases to expected cases) of 8.20 [4.09; 13.72], which was significantly different from 1. Among men, the ratio was 1.96 [0.79; 3.66], a ratio not significantly different from 1. A further investigation had to be conducted to confirm or refute the observed excess of cases, involving: - the identification and validation of all incident cancer cases (previously reported cases and new cases) during the study period, from 1995 to 2001, among the staff of the suspected building. - The establishment’s human resources department compiling lists of staff who worked in the suspected building during the study period. The study population consisted of all individuals who had worked for at least 6 months in the suspected building between January 1, 1995, and December 31, 2001, regardless of their administrative status, such as doctoral students, contract employees, tenured faculty, scholarship recipients, employees of the Ministry of Education, employees of the CNRS, etc. The 150 people who were working in the building in question in April 2002 had individual interviews with occupational health physicians. If the person had a history of cancer, the physician administered a medical questionnaire (to validate the cases), and a questionnaire describing their professional history was provided. The relatives of the six deceased individuals were contacted, and five families responded. One hundred twenty people were identified by administrative services as having worked in the building for at least six months and having since left. Only half of the addresses could be located, and approximately 40 completed questionnaires were returned to the physicians. Given the incomplete nature of the collected data, the analysis involved formulating several risk-maximizing hypotheses based on multiple population simulations (and thus variations in the number of validated cases). The study revealed the following: - Among women, according to the hypotheses: between 5 and 6 confirmed cases among 52 individuals, and a ratio ranging from 3.23 to 4.11, significantly different from 1. - Among men, according to the hypotheses: between 3 and 5 confirmed cases among 92 individuals, and a ratio ranging from 0.96 to 1.58, not different from 1. These results, which can only be considered indicative, seem to suggest an excess of breast cancer among women, findings consistent with recent data from the scientific literature showing an excess of cancer, particularly breast cancer, among research laboratory staff. The identification of risk factors for these cancers can only be achieved through multicenter epidemiological studies based on very large cohorts. In order to continue monitoring cancer incidence in this population, it is essential to have a list of these employees by name, year by year, including their first and last names, dates of birth, and gender, as well as a comprehensive record of all new cancer cases in this population.