Socially monogamous female birds routinely mate with males outside the pair bond. Three alternative hypotheses consider genetic benefits as the major driver behind the female strategy. The inbreeding avoidance hypothesis predicts that females paired with closely related males should seek copulations with distantly related extra-pair partners to avoid fitness loss from inbreeding depression; the outbreeding avoidance hypothesis predicts the opposite; the kin-selection hypothesis suggests that regardless of social mate relatedness, females should give related males extra-pair fertilization opportunities to gain inclusive fitness if the costs from inbreeding are minor. We test these hypotheses with a facultative cooperative breeder, the ground tit (Parus humilis). Social pairs of ground tits formed randomly with respect to genetic relatedness. In both bi-parental and cooperative groups, a female's engaging in extra-pair mating was independent of relatedness to her social mate; however, females preferred extra-pair sires to which they were more related than to their social mates. Moreover, females had higher relatedness with either their extra-group extra-pair sires in both bi-parental and cooperative groups, or within-group helper sires in cooperative groups, than expected by chance. When more than one potential extra-pair partner was available around a female's nest, she tended to select a relative. There was no indication of fitness reduction from extra-pair mating, which occurred at an intermediate level of inbreeding. These data support the kin-selection hypothesis, although there might be alternative nongenetic reasons associated with the extra-pair mating preference. Our finding offers a new explanation for why female birds pursue extra-pair mating. It also may broaden our understanding of the role of kin-selection in the evolution of cooperative society.