The prospect for improvement of aquaculture stocks by utilizing additive genetic variance, documented in several aquatic species during the last 15-20 years, should be exploited through long term selective breeding programmes to develop high performing breeds. Crossbreeding to capitalize on non-additive genetic effects, polyploidy or sex manipulation, may be applied for further improvement. The high fecundity facilitates concentration of resources in one or few breeding programmes in aquatic species. Genetic gain obtained in the nucleus may be disseminated throughout the industry with a minimum delay through multipliers. The high fecundity makes it possible to have a broad long term breeding objective in the nucleus while applying selection for a single or a few traits when producing growout animals. High benefit/cost ratios are therefore to be expected in breeding programmes for aquatic species. The two greatest obstacles of efficient family selection programmes are the present tagging system for which family groups have to be reared separately until individuals are large enough to be tagged and the lack of methodology to record traits on live breeding candidates. Microsatellite DNA profiling for family identification represents a major advance in selective breeding of aquatic species allowing different family groups to be kept in a common tank from fertilization onwards.