The prevalent conviction among ecologists that natural communities represent important and meaningful assemblages of organisms has prompted a diverse series of analyses. The present paper is an examination of some of the analyses as applied primarily to soil arthropods, with an explanation of some of the difficulties that have been encountered. The thesis contained herein is that the numerical abundance and spatial distribution of all species found must be taken into account before an understanding of community organization can be obtained. Analyses and definitions of the community have been used to approach the subject from several more or less distinct points of view-function, location, and biotic composition. Most studies have dealt with only one of these aspects, but some assumption has usually been made about the others. Macfadyen (1957) has recently reviewed the various community definitions that have been offered. In the present paper, particular attention is paid to the aspects of community structure that can be discovered from the occurrence and relative abundance of species, usually confined to a taxonomic group not larger than a phylum. The species composition of communities has been used as an approach to community analysis for at least 50 years. The data have been viewed from the standpoints of species frequency, species per unit of area, the spatial distribution of individuals, and the numerical abundance of species. In all of these, attempts have been made, graphically and mathematically to express some aspect of community structure.