The rare Australian venomous elapid snake ‘Echiopsis’ atriceps has been the subject of considerable taxonomic instability with the five known specimens assigned to four genera by various authorities. Phylogenetic affinities of the rare Elapognathus minor also are poorly understood and have been the subject of some disagreement. To examine the phylogenetic affinities of these two rare taxa, a molecular data set comprising 1680 base pairs of mtDNA was assembled from a representative of each of the terrestrial Australian viviparous elapid genera and two species of Drysdalia, a genus about which there also has been phylogenetic controversy. A total of 936 base pairs of 12S rRNA, 454 base pairs of 16S rRNA and 290 base pairs of cytochrome b mtDNA were sequenced for 15 species. The Asian elapid Naja naja was used as the outgroup. These mtDNA regions provided 195, 38 and 72 parsimony informative sites, respectively, for a total of 315 parsimony informative characters. Unweighted phylogenetic analyses were performed under both parsimony and neighbour-joining criteria. Parsimony analyses of the unweighted, combined data set resulted in a single fully resolved most parsimonious tree 1225 steps long. The neighbour-joining tree differed by only a single weakly supported branch. These data strongly support a sister group relationship between ‘Echiopsis’ atriceps and the Australian broadheaded snakes of the genus Hoplocephalus with a bootstrap value of 99%. Templeton tests soundly reject all previous taxonomic arrangements for this species. Our data also strongly support a sister group relationship between Elapognathus minor and Drysdalia coronata with a bootstrap value of 98%. Importantly, Drysdalia coronata and Drysdalia coronoides do not form a monophyletic group, supporting some previous studies. Based on our results, we allocate ‘Echiopsis’ atriceps to a new monotypic genus and re-describe Elapognathus to include ‘Drysdalia’ coronata.