Southwestern Australia (SWA) is a region of temperate Mediterranean climate isolated by desert from the rest of Australia. Since the Jurassic it has been a geologically stable area that resisted Cenozoic glaciations and today represents an ancient landscape characterized by subdued topography and nutrient-poor soils. Despite these ecological conditions, SWA contains an incredibly rich flora and fauna that includes a great diversity of endemic species and it recently has been identified as a biodiversity hotspot of international significance. Since the early recognition of the high floral diversity in SWA and subsequent recognition of high faunal diversity, much discussion has focused on the origins of this rich endemic biota. Two alternative models have been proposed--the Multiple Invasion Hypothesis and the Endemic Speciation Hypothesis. Multiple tests of these models have variously supported either one, but many of the tests have been poor. Here we use a phylogeny for the myobatrachid frog genus Heleioporus to distinguish between these hypotheses. Heleioporus comprises six species: five endemic to southwestern Australia with one from eastern Australia. A molecular phylogeny using two mitochondrial genes (ND2 and 12S rDNA) and one nuclear gene (rag1) was used to test alternative theories about the biogeography and the origin of diversity in this genus. Using a relaxed molecular clock, the divergence between the eastern and western species was dated at 25.60 M years, which is considerably older than previously suggested. Our phylogeny of Heleioporus is inconsistent with previous biogeographic hypotheses involving repeated invasions from the east to the west and some previous in situ models and instead strongly supports an ancient endemic speciation model. While the split between east and west appears to be contemporaneous with similar splits in Geocrinia (Anura) and Banksia (Proteaceae) it is much older than splits in a range of other taxa including other anurans.