The idea that extrapair paternity (EPP) in birds is part of a mixed reproductive strategy driven primarily by females is controversial. In cooperatively breeding American crows, we compared predictions of four female benefits hypotheses-the genetic diversity, good genes, genetic compatibility, and direct benefits hypotheses-to our predictions if EPP was primarily male driven. We found that genetically diverse broods were not more successful, extrapair young were not in better condition and did not have a higher survival probability, and, contrary to prediction, offspring sired by within-group extrapair males were more inbred than within-pair offspring. There was evidence of direct benefits, however: provisioning rate and number of surviving offspring were higher in groups containing within-group extrapair sires. Females therefore derived no apparent benefits from extragroup extrapair males but both direct benefits and genetic costs from within-group extrapair males. We suggest that males and females both influence the distribution of EPP in this system.