This paper reviews the current genetic knowledge on issues related with the efficient use of bucks in artificial insemination (AI). Differences between lines have been found to be relevant in semen production and quality traits not necessarily related to their specialisation as maternal or paternal lines. Accurate heritability estimates indicate that genetic selection to increase semen production by improving male libido and/or reducing the number of rejected ejaculates may not be effective. However, total sperm produced per ejaculate appears to be an interesting trait to select for. Semen pH has shown low to medium heritability estimates and a low coefficient of variation, therefore it is not advisable to attempt improvement by direct selection. In general, sperm motility traits have shown low heritabilities, but the rate of motile sperms per ejaculate has been considered convenient to select for. Morphological characteristics of the spermatozoa have been revealed as medium to highly heritable. There is evidence of high genetic correlations between sperm traits before and after freezing-thawing. There are few studies regarding the estimation of heterosis of seminal traits, but results indicate important and favourable direct and maternal heterosis in crosses between maternal lines. However, this has not been confirmed in a cross between 2 paternal lines. Until now, attempts to find parametric or non-parametric functions to predict ejaculate fertility through seminal characteristics routinely recorded in evaluations have been very unsatisfactory. Hence, it may be necessary to find other semen quality markers or assess some of those currently used in a more precise manner, or closer to the AI time, in order to improve the ability to predict ejaculate fertility. Several seminal characteristics phenotypically correlated to male fertility could be considered as potential traits to select for in order to genetically improve this trait. However, only the semen pH has been checked for this purpose, and results show a low genetic correlation of this trait with male fertility. Other traits can be studied in the future, but bearing in mind that the required experiments will need a large number of bucks for an accurate estimation of the genetic correlation of the trait with male fertility. This means that these experiments will be expensive and difficult to set up. The most common criterion for selecting paternal lines, average daily gain, seems not to be genetically correlated to male fertility and seminal traits. Thus, selection for average daily gain has no detrimental consequences on these traits, and a multi-trait selection, including growth rate and seminal traits directly related to an efficient AI semen dose production, is feasible in paternal lines. The male contribution to fertility after natural mating and after AI with semen doses with high concentration is negligible, although it has been found that under more restrictive AI conditions male contributions to fertility and litter size are low, but high...