In the hominin fossil record, pelvic remains are sparse and are difficult to attribute taxonomically when they are not directly associated with craniodental material. Here we describe the pelvic remains from the Dinaledi Chamber in the Rising Star cave system, Cradle of Humankind, South Africa, which has produced hominin fossils of a new species, Homo naledi. Though this species has been attributed to Homo based on cranial and lower limb morphology, the morphology of some of the fragmentary pelvic remains recovered align more closely with specimens attributed to the species Australopithecus afar-ensis and Australopithecus africanus than they do with those of most (but not all) known species of the genus Homo. As with A. afarensis and A. africanus, H. naledi appears to have had marked lateral iliac flare and either a weakly developed or non-existent acetabulocristal buttress or a distinct, albeit weakly developed, acetabulospinous buttress. At the same time, H. naledi has robust superior pubic and ischiopubic rami and a short ischium with a narrow tuberoacetabular sulcus, similar to those found in modern humans. The fragmentary nature of the Dinaledi pelvic assemblage makes the attribution of sex and developmental age to individual specimens difficult, which in turn diminishes our ability to identify the number of individuals represented in the assemblage. At present, we can only confidently say that the pelvic fossils from Rising Star represent at least four individuals based on the presence of four overlapping right ischial fossils (whereas a minimum of 15 individuals can be identified from the Dinaledi dental assemblage). A primitive, early Australopithecus-like false pelvis combined with a derived Homo-like true pelvis is morphologically consistent with evidence from the lower ribcage and proximal femur of H. naledi. The overall similarity of H. naledi ilia to those of australopiths supports the inference, drawn from the observation of primitive pelvic morphology in the extinct species Homo floresiensis, that there is substantial variation in pelvic form within the genus Homo. In the light of these findings, we urge caution in making taxonomic attributions-even at the genus level-of isolated fossil ossa coxae. (C) 2017 Elsevier Ltd. All rights reserved.