A monoclonal antibody that recognizes a nonphosphorylated epitope on the 168 kDa and 200 kDa subunits of neurofilament proteins has been used in an immunohistochemical study of cynomolgus monkey (Macaca fascicularis) and human neocortex. This antibody, SMI-32, primarily labels the cell body and dendrites of a subset of pyramidal neurons in both species. A greater proportion of neocortical pyramidal neurons were SMI-32 immunoreactive (ir) in the human than in the monkey. SMI-32-ir neurons exhibited consistent differences in the intensity of their immunoreactivity that correlated with cell size. The cellular specificity of SMI-32 immunoreactivity suggests that a subpopulation of neurons can be distinguished on the basis of differences in the molecular characteristics of basic cytoskeletal elements such as neurofilament proteins. The size, density, and laminar distribution of SMI-32-ir neurons differed substantially across neocortical areas within each species and between species. Differences across cortical areas were particularly striking in the monkey. For example, the anterior parainsular cortex had a substantial population of large SMI-32-ir neurons in layer V and a near absence of any immunoreactive neurons in the supragranular layers. This contrasted with the cortical area located more laterally on the superior temporal gyrus, where layers III and V contained substantial populations of large SMI-32-ir neurons. Both areas differed significantly from the posterior inferior temporal gyrus, which was distinguished by a bimodal distribution of large SMI-32-ir neurons in layer III. Differences across human areas were less obvious because of the increase in the number of SMI-32-ir neurons. Perhaps the most notable differences across human areas resulted from shifts in the density of the larger SMI-32-ir neurons in deep layer III. A comparison between the species revealed that isocortical areas exhibited greater differences in their representation of SMI-32-ir neurons than primary sensory or transitional cortical areas. A comparison of distribution patterns of SMI-32-ir neurons across monkey cortical areas and data available on the laminar organization of cortical efferent neurons suggests that a common anatomic characteristic of this chemically identified subpopulation of neurons is that they have a distant axonal projection. Such correlations of cell biological characteristics with specific elements of cortical circuitry will further our understanding of the molecular and cellular properties that are critically linked to a given neuron's role in cortical structure and function.