No fossil evidence has yet been found to establish affinities of the family. Comparative morphological studies based largely on genital characteristics (Dirsh 1965) and wing venation (Smart 1953; Ragge 1955) suggest that pneumorids are a primitive family derived from general Caeliferan stock even earlier than Locustopsidae and earlier than other living Caeliferan superfamilies (Ragge 1955). Rehn (1941) considered them an ancient group equivalent to the Tetrigoidea. Pneumorids resemble Acridoids in many respects and are clearly related to these, but the exact nature of the relationship is uncertain. They may be primitive Orthoptera which developed in parallel with the Acridoidea and share characters by convergence, or highly specialized Acridids which have become secondarily simplified (Dirsh 1965). There have been no morphology-based cladistic tests of these hypotheses. Molecular data of Flook and Rowell (1997) suggest that the Pneumoridae branch off after the Eumastacoidea, but before the Acridoidea.
Alexander and van Staaden (1989) suggested that north eastern forest species are basal and south western desert species derived, but no complete phylogeny is yet been available for the family. A molecular phylogeny is under construction. In the most recent systematic revision, Dirsh (1965) recognized 9 genera containing 17 species. However several genera and species are known from only one sex or a single individual and are of dubious status. Genera with non-inflated males are particularly troublesome. Dirsh considered them "parallel to the species with inflated bodies", placing them in separate genera (Pneumoracris, Paraphysemacris, Parabullacris) and largely discounting van Son's view (in litt.) that these were neotenic forms. Present data suggest that neither view is correct. A non-inflated male morph occurs alongside the inflated male in at least two species. Mating behavior, food plant specificity and morphological measurements clearly indicate that these are alternative morphs rather than simply neotenic forms (Alexander and van Staaden 1989). Consideration of food plants and geographic distribution suggests that the non-inflated genera may be alternative males of inflated species (i.e. Pneumoracris browni = Perinqueyacris namaqua; Paraphysemacris spinosus = Physemacris variolosus). However, the situation of Parabullacris vansoni with respect to the B. unicolor complex is less clear.