Today, triploidy is the most common method used to genetically improve marine bivalve aquaculture production. In the Pacific oyster Crassostrea gigas, triploidy affects many traits. The most proeminent one is reproductive effort, which is greatly reduced compared to diploids. The combined effects of physiological trade-offs between survival, growth and reproduction together with the reduction of reproductive effort in triploids explains the global enhancement of yield observed. However, the sterility of triploid oysters is only partial and some triploid individuals can show normal gametogenesis (i.e. similar to diploids), especially under warm conditions. In diploid oysters, the heritability of reproductive effort has been shown to be relatively low, but a clear relationship between survival during summer (a period of extreme mortality), a trait with high heritability, and allocation to reproduction has been demonstrated. Three specific issues related to triploidy can be identified. Firstly, as demonstrated in various species from crop to invertebrates, triploidization is known to directly affects the average performance (i.e. phenotypic mean), regardless of the breeders genetic values. Indeed, depending on the considered trait and eventually the age of the individuals, triploid Pacific oysters show lower (e.g. reproductive effort), or higher (e.g. growth) mean performances compared to diploids. Secondly, triploidy interacts with the transmission of additive genetic values between parents and their progeny. This is due to the unbalanced contribution of male and female parents, as demonstrated in triploid rainbow trout triplo in which 2/3 of the triploid genome comes from the dam, and 1/3 from the sire. Third, triploidization can increase non-additive genetic effects (notably due to the interaction between three alleles at each locus), as shown for flowering time in Brassica napus. All these ploidy-related characteristics strongly influence the possibility to select diploids on their breeding value to obtain improved triploid progenies, notably depending on the relative importance of additive and non-additive genetic variance for the selected traits. In order to further examine (1) the genetic basis of the observed variation in reproductive effort, both in diploid and triploid oysters, and (2) the relationship between reproductive effort in diploid and triploid sibs, a quantitative genetics experiment was set up. A full-factorial cross was performed by mating 2 sets of 16 males (16 diploids and 16 tetraploids) with 6 diploid females. This generated 3 experimental groups comprising of 96 full-sib families: a diploid group and 2 triploid groups obtained either by crossing tetraploid and diploid broodstock (" natural " triploids: 3nn) or by second polar body retention (" chemically induced " triploids: 3nc). These were reared together (i.e. mixed-family approach) to avoid environmental bias. Reproductive effort (estimated by image analysis of histological cross sections) and growth performances were first measured on 300 individuals/group of 6 month-old spat, in parallel with parentage analysis assisted by microsatellite markers. Analyses are currently in progress. The results that will be presented deal with (1) the effects of the ploidy level (2n versus 3n) and of the method of triploid production ("natural" versus "chemically induced") on phenotypic means and variance for growth and reproduction; and (2) the variance between families and the resulting estimated genetic parameters (heritability, genetic correlations) for these two traits. Our results will be discussed in light of the trade-off between growth and reproduction and the opportunity of selective breeding of diploid broodstock to produce improved triploid progenies will be tackled.