Previous research on the evolution of vision in mammals has typically focused on the duplication and loss of genes encoding opsins, the proteins that make up one moiety of the retinal photopigment complex. These gains and losses are typically attributed to invading new photic niches, with gains associated with bright light and losses with dim light niches. Additionally, the adaptive significance of opsin spectral tuning (i.e., preferential absorption wavelength) has largely remained a mystery for terrestrial mammals. For example, for almost 25 years researchers have been aware that some mammals possess ultraviolet-sensitive opsins whereas others possess violet-sensitive homologs. While scientists have proposed various explanations for this variation, there have been no formal tests of these hypotheses. In this dissertation, I expand upon this previous work in two ways: 1) exploring the loss of non-opsin visual genes in mammals occupying dim-light habitats, and 2) testing hypotheses of the evolution of an opsin’s spectral tuning in terrestrial environments. I used genome- and PCR-generated DNA sequence data in a comparative phylogenetic framework, and concluded the following: 1) subterranean mammals have variable degrees of visual gene decay, especially to the bright light visual system, as a result of their underground lifestyle, 2) armadillos, sloths and anteaters likely descended from a subterranean ancestor, as evidenced by the loss of genes encoding their bright light visual system, and 3) the ultraviolet/violet opsin in mammals has evolved in response to an increase in eye length and light exposure during mammalian history.