For the past 20 years in Australian archaeology edge analysis has been offered as a viable alternative to the conventional artefact taxonomic systems currently available. Thus White and O'Connell, in A Prehistory of Australia, New Guinea and Sahul (White and O'Connell 1982:66-67) suggest that analyses of early Australian flake assemblages which take into account attributes of the retouched and altered edges of the tools are likely to be more informative than those which assume that tools should fall into clear morphological patterns. In expressing doubts about the adequacy and usefulness of stone tool classifications based on artefact morphology, White's concept of edge analysis has been influential in the development of alternative ways of analysing stone tool assemblages. These include White's own work on ethnoarchaeology (White and Thomas 1972; White et al. 1977) the analysis of reduction sequences (Hiscock 1982), and studies of residues, and usewear (Kamminga 1982; Fullagar 1982). Paradoxically edge analysis has also contributed to the further development of morphologically based systems of artefact classification. When characterising the implements in their stone tool assemblages many Australian archaeologists now include measurements of edge angle, length of retouched edge, height of retouch and descriptions of retouch types along with the usual measurements and descriptions of implement size and shape (Jones 1971; Lampert 1981; Schrire 1982). Without wishing to discuss or criticise these subsequent developments, it might be useful at this time to look again at the original ideas advanced in favour of edge analysis and the claims made for it. This paper then seeks to assess the advantages or otherwise of edge analysis over other typological systems using the flaked stone assemblages from sites excavated in the New Guinea Highlands by J.P. White as a test case.