The evidence obtained from a comparison of the results of behavioral and physiological studies of the macaque and the squirrel monkey seems to reflect the considerable differences between the color vision of Old World and New World monkeys. These data also give us insight into the mechanisms of normal and defective color vision in man. The visual systems of macaque monkeys and normal humans are nearly identical. In both these systems the outputs from the various cones are added together in one set of information channels (nonopponent cells) to indicate the whiteness-blackness of a light. This channel is also the principal determinant of the brightness of a light. In other channels (opponent cells) the output from one of the three cone types is subtracted from the output of one of the others, in various combinations, to signal the hue of the light. The ratio of opponent to nonopponent cell activity in response to a flash of light signals the saturation of the light. The opponent cells are very sensitive to small differences in wavelength, particularly in certain spectral regions, and provide the organism with information about color differences in different parts of the visual field. Squirrel monkeys and protanomalous humans, on the other hand, possess color vision which is both weak and deviant from the normal. The weak color vision can be attributed to a paucity of opponent cells, relative to the number in the macaque. The dominance of the 500-nanometer minimum in the relative saturation of the spectrum is attributable to the relatively few RG cells in the animal's (and presumably the protanomalous human being's) visual system, and to the low activity rate in the RG cells, due to the spectrally displaced photopigment present in the long-wavelength receptors.