The optic chiasm is one of the most popular models for studying axon guidance. Here axons make a key binary decision either to cross the midline to innervate the contralateral hemisphere or to remain uncrossed. In rodents, midline interactions between axons from the two eyes are critical for normal development, as early removal of one eye systematically disrupts hemispheric projections from the remaining eye, increasing the crossed projection at the expense of the uncrossed. This is similar to the abnormal decussation pattern seen in albinos. This pattern is markedly different in marsupials where early eye removal has no impact on projections from the remaining eye. These differences are related to the location of the uncrossed projection through the chiasm. In rodents these axons approach the midline whereas in marsupials they remain segregated laterally. We provide anatomical evidence in man suggesting that, unlike in rodents, uncrossed axons are confined laterally and do not mix in each hemi-chiasm, which is a pattern similar to that found in marsupials. Further, we demonstrate electrophysiologically, using visual cortical evoked potentials, that the failure of one eye to develop in man has no impact on the hemispheric projections from the remaining eye. These data demonstrate that the mechanisms regulating chiasmal development in man differ from those in rodents but may be similar to those in marsupials. We suggest that mouse models of the organization and development of the optic chiasm are not common to placental mammals in general.