Résumé: The aquaculture industry has increasingly aimed at improving economically important traits like growth, feed efficiency and resistance to infections. Artificial selection represents an important window of opportunity to significantly improve production. However, the pitfall is that selection will reduce genetic diversity and increase inbreeding in the farmed stocks. Genetic tools are very useful in this context as they provide accurate measures of genetic diversity together with many additional insights in the stock status and the selection process. In this study we assessed the level of genetic variability and relatedness over several generations of two lines of experimentally selected European sea bass (Dicentrarchus labrax L.). The first line was selected for growth over three generations and the second line for both high and low weight loss under a starvation regime over two generations. We used a genomic approach (2549 single nucleotide polymorphism markers derived from double digest restriction site associated DNA sequencing) in combination with eight microsatellites to estimate genetic variation, relatedness, effective population size and genetic differentiation across generations. Individual heterozygosity estimates indicated that the selected lines showed no significant reduction in diversity compared with wild populations. There was, however, a decreasing trend in allelic richness, suggesting the loss of low frequency alleles. We compared the estimates of effective population size from genetic markers with pedigree information and found good correspondence between methods. This study provides important insights in the genetic consequences of selective breeding and demonstrates the operational use of the latest genomic tools to estimate variability, inbreeding and at a later stage domestication and artificial selection.