Sex is expensive. A population of females that reproduce asexually should prima facie have twice the growth rate of an otherwise equivalent anisogamous sexual population lacking paternal care, or a population with modes of paternal care that can be co-opted by parthenogenetic females. The two leading theories for the maintenance of sex require either synergistic interactions between deleterious mutations, or antagonistic epistasis between beneficial mutations. Current evidence is equivocal as to whether the required levels of epistasis exist. Here I show that a third factor, differential male mating success (or, more generally, higher variance in male than in female fitness), can drastically reduce mutational load in sexual populations with or without any form of epistasis. Differential mating success has the further advantage of being ubiquitous, and is likely to have preceded or evolved concurrently with anisogamy.