Previous work on the proposed Karnic subgroup, comprising languages formerly spoken in the Lake Eyre Basin, a large area of eastern-central Australia, has been reviewed and found to be wanting in several respects. Application of a modified lexicostatistic method, together with grammatical comparisons, suggests that the western branch of the proposed three-branch subgroup (the southern portion of which is excluded from the subgroup in one of the studies) may in fact not be genetically closely related to the central and eastern branches. Detailed reconstruction is still needed to confirm the genetic unity of this branch. Two languages formerly included in the eastern branch are shown to be not closely related. Recent analyses of the central branch are supported.