Provisioning behavior in altricial birds is often used to measure parental investment and is assumed to have fitness consequences to the parents providing it, with the benefits outweighing the costs. Here we investigate the fitness costs and benefits (parent survival and offspring recruitment) of provisioning behavior in wild house sparrows Passer domesticus, using long-term data from a pedigreed isolated population. We disentangled the long-term fitness consequences in terms of number of recruits, of provisioning behavior from those of other parental investments and individual quality through a cross-foster design. We accounted for extra-pair offspring in all analyses. Provisioning behavior confers social fitness benefits in terms of the number of recruits to both parents. Only in females we detected an influence individual quality: female sparrows with high provisioning frequencies were associated with more genetic recruits than those who provided food less frequently to their young, even though foster parents reared the offspring. We detected a relationship between annual survival probability and provisioning behavior only in males, but not in females. This finding, together with indirect benefits differing by sex, indicates that different selection pressures are acting on the sexes. Our study can show that it is justified to use provisioning behavior as a form of parental investment sensu Trivers, since we show that this behavior is costly to parents and that the genetic fitness benefits exceed the costs.