Questions: How can a resemblance (similarity or dissimilarity) measure be formulated to include information on both the evolutionary relationships and abundances of organisms, and how does it compare to measures lacking such information? Methods: We extend the family of Phylogenetic Diversity (PD) measures to include a generalized method for calculating pair-wise resemblance of ecological assemblages. Building on previous work, we calculate the matching/mismatching components of the 2 × 2 contingency table so as to incorporate information on both phylogeny and abundance. We refer to the class of measures so defined as "PD resemblance" and use the term "SD resemblance" for the traditional class of measures based on species diversity alone. As an illustration, we employ data on the diversity and stem density of shrubs of Toohey Forest, Australia, to compare PD resemblance to its SD resemblance equivalent for both incidence and abundance data. Results: While highly correlated, PD resemblance consistently measures assemblages as more similar than does SD resemblance, and tends to "smooth out" the otherwise skewed and truncated distribution of pair-wise resemblance indices of our high-turnover data set, resulting in nMDS ordinations with lower stress. Randomization of species distributions across assemblages indicates that phylogeny has made a significant contribution to the ordination pattern. Conclusions: PD resemblance measures, in addition to providing an evolutionary perspective, have great potential to improve distance-based analyses of community patterns, particularly if species responses to ecological gradients are unimodal and phylogenetically conserved.