Padian and Hutchinson (1997) provided a stem-based definition for Allosauridae, using Allosaurus and Sinraptor as specifiers, although mentioning that relationships among the various genera composing Allosauroidea was uncertain. To better contain this uncertainty, Sereno (1998) provided a first-order revision that included several other allosauroid genera as external specifiers that help to limit potential membership. Sereno (1998) did the same for two other clades within Allosauroidea (Sinraptoridae, Carcharodontosauridae) that are most often used in the literature. Padian et al. (1999) and Holtz (2000) did not cite this revision and repeated their earlier definition.

The active definition takes a middle road. Clearly, if relationships among the genera are as uncertain as Padian et al. (1999) described, then it would be better to exclude at least the specifying species of the two other allosauroid subclades (Sinraptor dongi, Carcharodontosaurus saharicus), leaving other species to fall where they will. In a recent summary (Holtz et al. 2004), the family (which is not defined) is shown as including only two genera (Allosaurus, Saurophaganax), but the relationships among several others within Allosauroidea (Carnosauria) is uncertain. To eliminate the possibility that coelurosaurians would be included (in the event that any of these specifiers is closer to Coelurosauria than to themselves—i.e., if Allosauroidea is not monophyletic), one additional external specifier is used (Passer domesticus). This is a first-order revision of the definition originally given by Padian et al. (1997).