The cladistic analysis of ectatommine ants by Lattke ( 1994) is reanalyzed and discussed. It is argued that Lattke's sample of taxa (in particular his choice of outgroups) is problematic because of the acceptance of unreliable previous analyses of ant subfamilial relationships. Additionally, Lattke's proposed reclassification is inconsistent cladistically. A revised matrix is presented with the addition of three outgroups and twelve characters ( 19 taxa and 41 characters in total). This matrix yields a single most parsimonious cladogram (L=116; CI=46; RI=61) with Ponerinae paraphyletic with respect to the other poneroid exemplars included (Apomyrma, Cerapachys and Cheliomyrmex). This result illustrates the need for a full taxonomic revision of the poneroid group.