In a seminal paper, Hammerstein and Parker (1987) described how sex roles in mate searching can be frequency dependent: the need for one sex to perform mate searching is diminished when the opposite sex takes on the greater searching effort. Intriguingly, this predicts that females are just as likely to search as males, despite a higher potential reproductive rate by the latter sex. This prediction, however, is not supported by data: male mate searching prevails in nature. Counterexamples also exist in the empirical literature. Depending on the taxon studied, female mate searching can arise in either low- or high-density conditions, and suggested explanations differ accordingly. We examine these puzzling observations by building two models (with and without sperm competition). When sperm competition is explicitly included, male mate searching becomes the dominant pattern; when it is excluded, male mate searching predominates only if we assume that costs of searching are higher for females. Consequently, two hypotheses emerge from our models. The multiple-mating hypothesis explains male searching on the basis of the ubiquity of sperm competition, and predicts that female searching can arise in low-density situations in which sperm can become limiting. It can also explain cases of female pheromone production, where males pay the majority of search costs. The sex-specific cost hypothesis predicts the opposite pattern of female searching in high-density conditions, and it potentially applies to some species in which sperm limitation is unlikely.