The electroretinogram (ERG) is an essential measure of retinal function for studying mouse models of retinal disease. Ketamine, in combination with xylazine and/or acepromazine, is the most commonly used anesthetic agent. Although it works well in most situations, some fragile mouse strains have high mortality rates with this ketamine cocktail. We compared isoflurane with the ketamine cocktail in a longitudinal study of light-adapted and dark-adapted ERGs in C57BL/6J mice. Waveforms were averaged, oscillatory potentials (OPs) were extracted by digital filtration, and key ERG parameters were analyzed. The ERG waveforms were qualitatively similar with both anesthetics, and the male and female ERG parameters did not show significant differences. For light-adapted ERGs, b-wave amplitude and implicit time, and wavelet index were decreased under isoflurane anesthesia, whereas for dark-adapted ERGs, a- and b-wave implicit times were decreased and wavelet index was increased. The dark-adapted b-wave amplitude showed a significant inverse correlation with animal weight and age. Rod phototransduction gain and the Naka-Rushton n and R (max) parameters were the same for both anesthetics, and only the Naka-Rushton log k parameter was significantly elevated for isoflurane anesthesia. We propose that isoflurane is a satisfactory alternative to the ketamine cocktail for anesthesia in the mouse ERG. Precise quantitative comparisons, however, should only employ study designs using isoflurane versus isoflurane, or ketamine versus ketamine. Moreover, in light of the effects of both isoflurane and the ketamine cocktail on blood glucose levels, it would be prudent to control the fasting state of the animals in quantitative ERG studies.