Mating is rarely random in nature, but the effects of mate selection on offspring performance are still poorly understood, even in well-established models such as salmonid fish. We sampled wild lake char (Salvelinus umbla) and used their gametes to investigate the genetic consequences of different mating scenarios. In a first study, we used full-factorial breeding to experimentally separate additive genetic from maternal environmental effects. This led to 60 families and in total 1,073 embryos that were raised singly after sublethal exposures to the pathogen Aeromonas salmonicida, the common pollutant ethinylestradiol, or water only. Contrary to predictions of ‘good genes’ sexual selection, offspring of more yellow males were less tolerant to the pathogen than offspring of pale males, while male coloration did not predict offspring tolerance to ethinylestradiol. However, increased kinship between the parents had strong negative effects on embryo development in all treatment groups. In a second experiment, we monitored 1,464 singly-raised embryos that resulted from 70 pair-wise sperm competition trials. These embryos were raised in an environment that supports the growth of symbiotic microbes (sublethal stress) or in a clean environment. Offspring of yellow males were again less stress tolerant than those of pale males, and embryo development was again slowed down with increasing genetic similarity between the parents. We conclude that genetic benefits of mate selection would be strongest if females avoided genetic similarity during mate selection, for example based on MHC-linked signals, while male breeding colors seem more relevant in intra-sexual selection.