Most butter£y species can be characterized as capital breeders, meaning that reproductive output is strongly coupled to the amount of resources they have procured during the larval stage. Accordingly, female fecundity is generally correlated with female mass, both within and across species. However, the females of some species can be partly characterized as income breeders, in the sense that their reproductive output is dependent not only on larval-derived capital but also on resources acquired during the adult stage. These adult resources can be derived from female feeding or from male-transferred nuptial gifts. Recent studies on the within-species e¡ects of multiple matings on female ¢tness show that females generally gain directly from multiple matings in terms of increased lifetime o¡spring production. Here, we test whether the positive e¡ects of multiple mating on female ¢tness also hold at a comparative level, by conducting a laboratory study of female reproductive output in eight pierid species that di¡er in lifetime female mating frequency. Female reproductive output, measured as cumulative egg mass divided by female mass, increased signi¢cantly with polyandry (r 0.942, p 5 0.001), demonstrating that the positive e¡ect of mating rate on female reproductive ¢tness also holds between species. The positive e¡ect of male nutrient contribution is substantial, and the per capita reproductive output is more than twice as high in the most polyandrous species as in the most monandrous ones. Hence, the positive net e¡ect of the ejaculates is highly substantial, although males and females can have sexual interests that run counter to each other, setting the stage for sexually antagonistic coevolution, so that the various component parts of the male ejaculateösperm, nutrients, anti-aphrodisiacs, and gonadotrophic hormonesömay each correspond to a separate con£ict^cooperation balance between the sexes. Two scenarios for the evolution of nuptial gifts in butter£ies are discussed, one arguing that variation in larval food is the underlying factor and the other arguing that sexually antagonistic coevolution is the driving force. The two views are complementary rather than mutually exclusive, although the former hypothesis predicts that variation in female mating rate depends on variation in larval food availability, whereas the latter suggests that variation in female mating rate between species results from species-speci¢c idiosyncrasies.