The question of whether the large area occupied by the primate fovea in the visual cortex (V1) is the result of a selective amplification of the central visual field, or whether it merely reflects the ganglion cell density of the retina, has been a subject of debate for many years. Measurements of the ganglion cell densities are made difficult by lateral displacements of cells around the fovea and the occurrence of amacrine cells in the ganglion cell layer. We have now identified and counted these amacrine cells by GABA immunocytochemistry and by retrograde degeneration of ganglion cells. By reconstructing the fovea from vertical and horizontal serial sections, we were able to measure the densities of cones, cone pedicles and ganglion cells within the same retina. We found 3-4 ganglion cells for every foveal cone. This ratio decreased to one ganglion cell per cone at an eccentricity of 15-20 deg (3-4 mm) and in peripheral retina there are more cones than ganglion cells. The ganglion cell density changes by a factor of 1000-4000 between peripheral and central retina. A comparable gradient has been reported for the representation of the peripheral and central visual field in V1. We suggest that ganglion cell density can fully account for the cortical magnification factor and there is no need to postulate a selective amplification of the foveal representation.