The onychopod ‘cladoceran’ genus Bythotrephes Leydig is reviewed and redescribed based on introduced material of the B. cederstroemi‘form’ from the Great Lakes of North America. Reviewed briefly are the systematic history, introduction into North America, and natural history of the genus. Using S.E.M. and light microscopy, we then illustrate a mixture of plesiomorphic and (predominantly) derived features, most of which can be explained as modifications for predation. An emended generic diagnosis includes unique and previously poorly known characters that serve to distinguish Bythotrephes from the confamilial genus Cercopagis, such as a distinctly bilobed labrum, spinose anterior mandibular process, small (exopodal?) spines on the external face of thoracopods 1–3, and a gnathobasic process (proximal endite) on thoracopod 1. A redescription, combined with observations on the external morphology, highlights several morphological peculiarities. These include the strongly bilobed labrum, the loss of one of the original two pairs of maxillae (with a question raised as to which pair has been lost), an unusual point of origin for the ‘proximal endite’ (termed herein the ‘gnathobasic process’ because of uncertainty of homology) on thoracopods two and three, a bulbous process of unknown significance located just posterior to the last thoracopod, and the presence of what might be a remnant of the food groove believed to have been present in a hypothesized ancestral group leading to the Onychopoda. The problem of whether Bythotrephes contains one dimorphic species or two species, B. Iongimanus and B. cederstroemi, is reviewed briefly, and the morphological feature in which these forms differ (the caudal process) is described in detail for the cederstroemi form. Comparison with descriptions of other cercopagidids suggests that Bythotrephes is a highly derived taxon, with short‐abdomen genera of the Podonidae and Polyphemidae (e.g., Polyphemus) appearing closer to the onychopod phyletic stem, and with the more closely related genus Cercopagis more derived still. Phylogeny within the Cladocera, and the possible origin of the Cladocera from a cyclestheria‐like ancestor among the spinicaudate conchostracans, is reconsidered. A hypothesis of evolutionary relationships is presented that depicts the ctenopods and anomopods as having arisen from a hypothetical Cyclestheria‐like ancestor, with the predatory cladoceran taxa Haplopoda and Onychopoda being derived from the base of the anomopod lineage. Some of the many problems in the suggested phylogeny are discussed.