Wolbachia are capable of eliciting a variety of reproductive phenotypes from their hosts, including the production of an all-female progeny through embryonic male-killing. To date, phylogenetic analyses indicate six independent acquisitions of the ability to kill male embryos among Wolbachia strains which infect insects. Of these six strains, only one appears to have experienced horizontal transmission between host species while maintaining a male-killing phenotype. The rarity of male-killing Wolbachia and their disjunct phylogenetic relationships is surprising, given the apparently common occurrence of horizontal transfer involving Wolbachia strains causing other phenotypes. A male-killing Wolbachia strain examined here in Drosophila borealis represents a second case of apparent horizontal transmission, based on its close relationship to a male-killing strain in a distantly related Drosophila species. The results reported here show that this Wolbachia has maintained a stable phenotype in D. borealis over a period of at least 50 years, and that a similar strain elicits the same male-killing phenotype in a second Drosophila species, indicating that male-killing may be a stable long-term strategy. Sampling bias and/or a lack of suitable hosts are discussed as possible causes of the low frequency of male-killers identified among Wolbachia strains.