Résumé
Common carp was the first domesticated fish species and is now a major species in worldwide aquaculture. Although research has generated a considerable body of knowledge about its genetics, genetic improvement programmes are virtually non-existent in this species and only crossbreeding is used in practice. The main reasons are the inefficiency of methods usually effective in other aquaculture species (monosexing, triploidy), and the bad image of selective breeding which is difficult to implement in ponds, and for which reliable genetic parameters were missing until recently. These parameters are now becoming available through the use of parentage assignment with microsatellite markers. However, the question of selective breeding for growth remains far from simple, as growth is essentially linked to the productivity of the pond and to the stocking density of fish. We conclude that breeding should focus primarily on quality traits and possibly disease resistance which are easier to assess. Growth is a long term objective which can only be evaluated with large cumulated gains. Transgenesis could be an option for this goal, but along with the biosafety and acceptability issues it raises, its present results are quite disappointing, and therefore this technique does not at present seem to be an immediate answer to the situation.