We have used antibodies specific to either the red/green-or blue- sensitive cones in order to compare their ratio and distributions to that of the rods in the retinae of 3 primate species that differ in their capacity for color vision. We have found that the monoclonal antibody CSA-1 (Johnson and Hageman, 1988) and the polyclonal antibody 4942A, specific to the red- and green-cone opsin (Lerea et al., 1989), applied to retinal whole-mounts labeled approximately 90% of all cones in the diurnal Old-World rhesus monkey (Macaca mulatta) and all of the cones in the nocturnal New-World owl monkey (Aotus trivirgatus) and nocturnal prosimian bushbaby (Galago garnetti). The polyclonal antibody 108B, specific to the blue-cone opsin (Lerea et al., 1989), labeled about 10% of the cones across the entire surface of the rhesus monkey retina, but failed to label any cones in the retina of the 2 nocturnal species. Only the retina of the rhesus monkey possessed an all-cone foveola in which the density of cone inner segments was 17-fold greater than that in the fovea of the owl monkey or bushbaby retina. Surprisingly, the density of cones per unit area outside of the fovea was comparable in all 3 species. Rod density in the dorsal retina was elevated in all animals examined, but was 2-3 times greater in the nocturnal species than in the rhesus monkey retina. Application of the photoreceptor-class-specific antibodies may provide further insights into the evolution and development of wavelength sensitivity in the retina, as well as enhance our understanding of normal and abnormal color vision in humans.