SummaryThis paper reports on a method of extending species descriptions by incorporating information from the taxonomic hierarchy into a numerical classification of plant community types based on species. The basic species data are expanded to include information on higher taxonomic categories, such as genus, family, order, or class. Each stand is described by the presence, or abundance, of taxonomic categories but, unlike previous methods, only the necessary levels of the hierarchy are included. “Necessary” is here defined to be the highest level in the taxonomic hierarchy which uniquely identifies some specific taxon in the data set. If only a single member of an order is present in the data, then only taxonomic categories equal to or higher than, order are necessary to uniquely identify that taxon. Information on family, genus or species is simply redundant. Where several members of the same genus are present then all levels of the hierarchy will be recorded since they are all necessary for unique identification. Having obtained stand descriptions in terms of the taxonomic categories then classification can proceed normally using standard techniques. The method was applied to a small set of data, for which a variety of previous results was already available for comparison. The method appears to provide an effective means of extending species descriptions.