AbstractThe distribution of divergence times between member species of a community reflects the pattern of species composition. In this study, we contrast the species composition of a community against the meta‐community, which we define as the species composition of a set of target communities. We regard the collection of species that comprise a community as a sample from the set of member species of the meta‐community, and interpret the pattern of the community species composition in terms of the type of species sampled from the meta‐community. A newly defined effective species sampling proportion explains the amount of the difference between the divergence time distributions of the community and that of the meta‐community, assuming random sampling. We propose a new index of phylogenetic skew (PS), as the ratio of the maximum‐likelihood estimate of the effective species sampling proportion to the observed sampling proportion. APSvalue of 1 is interpreted as random sampling. If the value is >1, the sampling is suspected to be phylogenetically skewed. If it is <1, systematic thinning of species is likely. Unlike other indices, thePSdoes not depend on species richness as long as the community has more than a few members of a species. Because it is possible to compare partially observed communities, the index may be effectively used in exploratory analysis to detect candidate communities with unique species compositions from a large number of communities.