Although raphitomid snails are a dominant component of gastropod communities in deep seas worldwide, their systematics is still largely tentative. We assembled the most complete sampling of Raphitomidae from south-eastern Australia to date. Based on morphological and molecular data from this material, we produced a robust phylogenetic framework and used it to delimit genera. For the focus area, our results show a large proportion of undescribed species- and genus-level taxa, 11 of which are formally described herein. We demonstrate that the examination of purely morphological characters rarely suffices for the purpose of accurate genus delimitation. As a result, some traditionally highly diverse raphitomid genera (such as Gymnobela) turn out to be artificial assemblages of several unrelated, mostly undescribed, genus-level lineages. Our data suggest that comparable configurations of shell and radular features, observed at the genus level, commonly do not reflect true phylogenetic relationships. However, our results are inconclusive as to whether homoplasy or conservatism are the drivers of this phenomenon. Accommodating for the inevitable sampling biases, south-eastern Australia appears as a possible hotspot for both raphitomid diversity and endemism, when compared with adjacent areas.