Many studies quantify how polyandry affects female fitness by allowing females to mate with one or several males. But even if the number of matings is standardised, such studies conflate any costs of interacting with males with potential benefits of receiving sperm from multiple mates, obscuring the benefits of polyandry. We conducted a 2x2 factorial experiment on the mosquitofish, Gambusia holbrooki to quantify the independent effects of male harassment and polyandry. We artificially inseminated virgin females with sperm from either one or five males. Females were housed in the presence or absence of reproductively-ablated males who could harass, but not mate with them. Our design ensured that the number of males inseminating a female was independent of the level of male harassment and the number of matings she received. Females not housed with males were instead housed with immature females to maintain densities across treatments. Unexpectedly, females that experienced sexual harassment were more likely to give birth, had shorter gestation periods and gave birth to larger broods. Furthermore, polyandrous females were more likely than monandrous females to give birth, and their sons reached maturity faster than those of monandrous females. We found no detectable costs to females of male harassment when the direct costs of mating were absent and that, in the absence of mating costs, there are direct and indirect fitness benefits of being inseminated by multiple males. If the costs of the act of mating are small or absent, polyandry will benefit female G.holbrooki producing their first brood.