Matrotrophy, the provisioning of embryos between fertilization and birth, creates the potential for conflict between mothers and embryos over the level of maternal investment. This conflict is predicted to drive the evolution of reproductive isolation between populations with different mating systems. In this study, we examine whether density-driven mating system differences explain the patterns of asymmetric reproductive isolation observed in previous studies involving four populations of the matrotrophic least killifish, Heterandria formosa. Minimum sire number reconstructions suggested that two populations characterized by low densities had lower levels of concurrent multiple paternity than two populations characterized by high densities. However, low levels of genetic variation in the low-density populations greatly reduced our probability of detecting multiple mating in them. Once we took the lower level of genetic variation into account in our estimations, high levels of multiple paternity appeared the rule in all four populations. In the population where we had the greatest power of detecting multiple mating, we found that multiple paternity in H. formosa typically involves multiple sires contributing to offspring within the same brood instead of different fathers contributing to distinct, simultaneously provisioned broods. Paternity was often skewed towards one sire. Our results suggest that differences between H. formosa populations in the levels of multiple paternity are not sufficient to explain the reproductive isolation seen in previous studies. We suggest that other influences on maternal–foetal conflict may contribute to the pattern of reproductive isolation observed previously. Alternatively, the asymmetric reproductive isolation seen in previous studies might reflect the disruption of maternal–foetal coadaptation.