Zooarchaeologists have established several criteria for differentiating hominid- and hyena-derived faunal assemblages. In some cases, however, the patterns of skeletal part representation and bone surface modification on which these criteria are ultimately based have been observed in fossil bone assemblages of unknown origin, rather than in modern assemblages of known origin. When the proposed criteria are evaluated within an actualistic framework, only three are able to differentiate between hominid- and hyena-created faunal assemblages. I suggest that only these three criteria—proportions of carnivores to ungulates in the assemblage, the preserved condition of long bone specimens (either as whole cylinders or as splintered shaft fragments), and the types of bone surface modifications—should be retained as important factors in a diagnosis of the ancient bone-collector. The remaining four criteria—the relative proportion of horn pieces in the assemblage, the relative representation of podial bones, the relative representation of small and large bovid skeletal parts, and bovid mortality profiles—are not relevant or applicable to the problem of differentiating hominid- from hyena-derived faunal assemblages. Copyright © 2002 John Wiley & Sons, Ltd.