Three cases of childhood cancer occurred between 1995 and 1999 at a preschool built on a former industrial brownfield site in Vincennes (Val-de-Marne). An initial assessment in 2000 had led to the recommendation that no further investigation be conducted, but the occurrence of a new case a year later prompted a reconsideration of the possibility that a local factor might be responsible for an increased risk of cancer. We studied cancer incidence in the cohort of children who attended the school, on the one hand, and among children residing in the southern neighborhood of Vincennes, on the other, during the periods 1990–1994 and 2000–2004, which flanked the alert period. Standardized incidence ratios (SIRs) were calculated using incidence rates from the retrospectively reconstructed Val-de-Marne childhood cancer registry as a reference. The excess that triggered the alert was time-limited, and there was no significant increase in incidence during the five-year period following the alert. The South neighborhood had no known risk factors and no specific exposure likely to cause a local excess of cancer. A local epidemiological investigation could not contribute to a better understanding of the causes of childhood cancers. In conclusion, there was no indication to extend specific epidemiological surveillance. (R.A.)