Captive-bred mallards Anas platyrhynchos have been released for hunting purposes at a very large scale in Europe since the mid-1970s. In spite of a potential genetic impact, the actual contribution of restocked mallards to the genome of the target population has received little attention. The genetic structure of modern wild mallards in the Camargue, Southern France, was assessed from two samples: one originating from shot birds in hunting bags and one from presumed wild ducks captured alive in a hunting-free reserve. Reference samples originated from five mallard farms, as well as from museum samples collected before the mid-1970s (i.e. before massive mallard releases started). Our results revealed that the genetic signature of wild wintering mallards has not changed significantly because museum and presumed wild samples from the Camargue hunting-free nature reserve were genetically similar, and clearly differentiated from the farm mallards. This suggests that mallard releases in the Camargue or elsewhere in France, although massive, have not actually translated into complete admixture of wild and captive genomes, most likely due to low survival of released birds once in the wild. Nevertheless, although genetic introgression of the wild population by captive-bred was contained, we found significant rates of hybridization between wild and captive-bred mallards in modern samples. This result suggests that long-term releases of captive-bred mallards, if carried on at such large scale, could compromise irreversibly the genetic structure and composition of European mallards. This work contributes to fill in the gap on the monitoring of the genetic consequences of large-scale game releases for exploitation.