Data generated from an earlier study have suggested a model in which greater numbers of long, vertical, associative axons may occur in the anterior cingulate cortex of schizophrenic patients relative to control subjects. This hypothesis has now been tested using neuron-specific antibodies raised against the 200-kilodalton neurofilament subunit, a component of neuronal cytoskeleton, to immunostain axons of human postmortem cingulate cortex. A manual method for counting axons in the region of layer II and sublamina IIIA has been designed and applied blindly to parallel control and schizophrenic immunoprocessed specimens. The results show that there are 25% more vertical axons in the schizophrenic than in the control specimens. Preferentially higher numbers of both long vertical axons (62%) and axons associated with blood vessels (52%) have also been noted in the schizophrenic specimens. By contrast, the number of large-caliber horizontal axons was the same in the two groups; therefore, the greater number of vertical axons in schizophrenic specimens does not appear to represent a nonspecific effect. When these data are corrected for the effects of several confounding variables using analysis of covariance, the overall pattern of the results persists. These findings suggest the possibility that there might be an increase of associative inputs into the anterior cingulate cortex of schizophrenic patients, although it is not clear at present whether the differences noted, if replicative, may be primarily or perhaps only secondarily related to the disorder.