Microspectrophotometric measurements were performed on 217 photoreceptors from cynomolgus, Macaca fascicularis, and rhesus, M. mulatta, monkeys. The distributions of cell types, for rods and blue, green, and red cones were: 52, 12, 47, and 44, respectively, for the cynomolgus, and 22, 4, 13, and 13 for the rhesus. Visual cells were obtained fresh (unfixed), mounted in various media (some containing 11-cis-retinal), and then located visually under dim red (650 nm) illumination. Absolute absorbance (A), linear dichroism (LD), and bleaching difference (BD) absorbance spectra were recorded through the sides of outer segments. The spectra were subjected to rigorous selection criteria, followed by digital averaging and Fourier transform filtering. Statistical methods were also applied to the accepted samples in the estimation of population means and variances. The wavelength of mean peak absorbance (lambda max) and the standard error at 95% certainty of the rod and blue, green, and red cone pigments in cynomolgus were 499.7 +/- 2.5, 431.4 +/- 4.2, 533.9 +/- 2.4, and 565.9 +/- 2.8 nm, respectively. The rhesus pigments were statistically indistinguishable from the cynomolgus, having lambda max of approximately 500, 431, 534, and 566 nm. Statistical tests did not reveal the presence of a lambda max subpopulation (i.e., anomalous pigments). The bandwidth of each alpha-band was determined in two segments, giving rise to the longwave half-density (LWHDBW), shortwave half-density (SWHDBW), and total half-density (THDBW) bandwidths. The LWHDBW was found to have the smallest variance. Both the LWHDBW and the THDBW showed linear dependence on the peak wavenumber (lambda max)-1 for the four macaque pigments.