Hypothesis: Species infected with male-killing endosymbionts may exhibit reversal in sexual selection, with males exhibiting a preference for uninfected females. Organism: Drosophila innubila, a species in which a substantial proportion of the females are infected with male-killing Wolbachia. Methods: Laboratory experiments in which males were allowed to choose between infected and uninfected females within a population cage. Virgin or non-virgin males and females were tested separately in all four combinations, as prior mating history may affect motivation to mate and thus choosiness. In a separate experiment, males were allowed to mate repeatedly and offspring production was monitored to determine if multiple mating adversely affects male fertility. Results: In the mate choice tests, the number of infected and uninfected females that mated was very similar in all tests, thus providing no evidence that males prefer uninfected females. Male fertility declined substantially in successive matings, indicating that they do not have unlimited capacity to sire offspring through repeated mating. New hypothesis: If infected and uninfected females differ in some quantitative phenotypic trait, the probability that a female is infected depends not only on her phenotype but also on the prevalence of infection among females in the population. Hence, the phenotypic cues that might be used as the basis for adaptive male mate choice are likely to be ambiguous, thus impeding the evolution of male choosiness.