The relative size of the hypoglossal canal has been proposed as a useful diagnostic tool for the identification of human-like speech capabilities in the hominid fossil record. Relatively large hypoglossal canals (standardized to oral cavity size) were observed in humans and assumed to correspond to relatively large hypoglossal nerves, the cranial nerve that controls motor function of the tongue. It was suggested that the human pattern of tongue motor innervation and associated speech potential are very different from those of African apes and australopithecines; the modern human condition apparently appeared by the time of Middle Pleistocene Homo. A broader interspecific analysis of hypoglossal canal size in primates conducted in 1999 has rejected this diagnostic and inferences based upon it. In an attempt to resolve these differences of opinion, which we believe are based in part on biased size-adjustments and/or unwarranted assumptions, a new data set was collected and analyzed from 298 extant hominoid skulls, including orangutans, gorillas, chimpanzees, bonobos, siamang, gibbons, and modern humans. Data on the absolute size of the hypoglossal nerve itself were also gathered from a small sample of humans and chimpanzee cadavers. A scale-free index of relative hypoglossal canal size (RHCS) was computed as 100 8 (hypoglossal canal area0.5/oral cavity volume0.333). No significant sexual dimorphism in RHCS was discovered in any species of living hominoid, but there are significant interspecific differences in both absolute and relative sizes of the hypoglossal canal. In absolute terms, humans possess significantly larger canals than any other species except gorillas, but there is considerable overlap with chimpanzees. Humans are also characterized by large values of RHCS, but gibbons possess an even larger average mean for this index; siamang and bonobos overlap appreciably with humans in RHCS. The value of RHCS in Australopithecus afarensis is well within both human and gibbon ranges, as are the indices computed for selected representatives of fossil Homo. Furthermore, the size of the hypoglossal nerve itself, expressed as the mass of nerve per millimeter of length, does not distinguish chimpanzees from modern humans. We conclude, therefore, that the relative size of the hypoglossal canal is neither a reliable nor sufficient predictor of human-like speech capabilities, and paleoanthropology still lacks a quantifiable, morphological diagnostic for when this capability finally emerged in the human career.