A new hypothesis for the evolution of Bilateria is presented. It is based on a reinterpretation of the morphological characters shared by protostomes and deuterostomes, which, when taken together with developmental processes shared by the two lineages, lead to the inescapable conclusion that the last common ancestor of Bilateria was complex. It possessed a head, a segmented trunk, and a tail. The segmented trunk was further divided into two sections. A dorsal brain innervated one or more sensory cells, which included photoreceptors. "Appendages" or outgrowths were present. The bilaterian ancestor also possessed serially repeated "segments" that were expressed ontogenetically as blocks of mesoderm or somites with adjoining fields of ectoderm or neuroectoderm. It displayed serially repeated gonads (gonocoels), each with a gonoduct and gonopore to the exterior, and serially repeated "coeloms" with connections to both the gut and the exterior (gill slits and pores). Podocytes, some of which were serially repeated in the trunk, formed sites of ultrafiltration. In addition, the bilaterian ancestor had unsegmented coeloms and a contractile blood vessel or "heart" formed by coelomic myoepithelial cells. These cells and their underlying basement membrane confine the hemocoelic fluid, or blood, in the connective tissue compartment. A possible scenario to account for this particular suite of characters is one in which a colony of organisms with a cnidarian grade of organization became individuated into a new entity with a bilaterian grade of organization. The transformation postulated encompassed three major transitions in the evolution of animals. These transitions included the origins of Metazoa, Eumetazoa, and Bilateria and involved the successive development of poriferan, cnidarian, and bilaterian grades of organization. Two models are presented for the sponge-to-cnidarian transition. In both models the loss of a flow-through pattern of water circulation in poriferans and the establishment of a single opening and epithelia sensu stricto in cnidarians are considered crucial events. In the model offered for the cnidarian-to-bilaterian transition, the last common ancestor of Eumetazoa is considered to have had a colonial, cnidarian-grade of organization. The ancestral cnidarian body plan would have been similar to that exhibited by pennatulacean anthozoans. It is postulated that a colonial organization could have provided a preadaptive framework for the evolution of the complex and modularized body plan of the triploblastic ancestor of Bilateria. Thus, one can explore the possibility that problematica such as ctenophores, the Ediacaran biota, archaeocyaths, and Yunnanozoon reflect the fact that complexity originated early and involved the evolution of a macroscopic compartmented ancestor. Bilaterian complexity can be understood in terms of Beklemishev "cycles" of duplication and colony individuation. Two such cycles appear to have transpired in the early evolution of Metazoa. The first gave rise to a multicellular organism with a sponge grade of organization and the second to the modularized ancestor of Bilateria. The latter episode may have been favored by the ecological conditions in the late Proterozoic. Whatever its cause, the individuation of a cnidarian-grade colony furnishes a possible explanation for the rapid diversification of bilaterians in the late Vendian and Cambrian. The creation of a complex yet versatile prototype, which could be rapidly modified by selection into a profusion of body plans, is postulated to have affected the timing, mode, and extent of the "Cambrian explosion." During the radiations, selective loss or simplification may have been as creative a force as innovation. Finally, colony individuation may have been a unique historical event that imprinted the development of bilaterians as the zootype and phylotypic stage. (ABSTRACT TRUNCATED)