Article dans une revue: Using a simple theoretical decision model and an original database, we were able to elicit the distribution of the utility value of having a child with Down syndrome for a large sample of French pregnant women (n = 28,341) between 2003 and 2007. We found that, on a scale where the value of a fetal death is 0 and the value of a healthy child is 1, the mean value for a child with Down syndrome is about −0.6. Assuming that the policymaker used the same decision model as the women, we infer from the French amniocentesis reimbursement regulation an implicit social value for a child with Down syndrome of −2.5. We conclude from our study that the policymaker is more likely to prevent the birth of children with Down syndrome than French women themselves.