Species of Tetillidae are distributed worldwide. However, some genera are unresolved and only a few genera and species of this family have been described from the Antarctic. The incorporation of 25 new COI and 18S sequences of Antarctic Tetillidae to those used recently for assessing the genera phylogeny, has allowed us to improve the resolution of some poorly resolved nodes and to confirm the monophyly of previously identified clades. Classical genera such as Craniella recovered their traditional diagnosis by moving the Antarctic Tetilla from Craniella, where they were placed in the previous family phylogeny, to Antarctotetilla gen. nov. The morphological re-examination of specimens used in the previous phylogeny and their comparison to the type material revealed misidentifications. The proposed monotypic new genus Levantinella had uncertain phylogenetic relationships depending on the gene partition used. Two more clades would require the inclusion of additional species to be formally established as new genera. The parsimony tree based on morphological characters and the secondary structure of the 18S (V4 region) almost completely matched the COI M1-M6 and the COI+18S concatenated phylogenies. Morphological synapomorphies have been identified for the genera proposed. New 15 28S (D3-D5) and 11 COI I3-M11 partitions were exclusively sequenced for the Antarctic species subset. Remarkably, species within the Antarctic genera Cinachyra (C. barbata and C. antarctica) and Antarctotetilla (A. leptoderma, A. grandis, and A. sagitta), which are clearly distinguishable morphologically, were not genetically differentiated with any of the markers assayed. Thus, as it has been reported for other Antarctic sponges, both the mitochondrial and nuclear partitions used did not differentiate species that were well characterized morphologically. Antarctic Tetillidae offers a rare example of genetically cryptic (with the traditional markers used for sponges), morphologically distinct species.