The electroretinogram (ERG) of the rhodopsin knockout (rho-/-) mouse of Humphries et al. (1997) (Humphries et al., 1997) was studied for evidence of light-evoked rod activity and to describe the cone function. The rho-/- retina develops normal numbers of rod and cone nuclei, but the rods have no outer segments, and no rhodopsin is found by immunohistochemistry. The dark-adapted ERG threshold was elevated 4.7 log units above wild-type (WT) control mice, indicating that any residual rod responses were reduced >50,000-fold, consistent with a complete functional knockout. The dark-adapted rho-/- ERG had a cone waveform, and the spectral sensitivity peaked near 510 nm for both dark-adapted and light-adapted conditions, without evidence of a Purkinje shift. The light-adapted ERG b-wave amplitude of young rho-/- mice was the same as WT. The amplitude remained steady up to postnatal day P47, but thereafter it declined to only 1-2% by P80 when no cone outer segments remained. Cone b-wave threshold of dark-adapted rho-/- mice was -1.07 +/- 0.39 log cd-s/m2 (n = 17), which is 1.27 log units more sensitive than light-adapted thresholds against a rod-suppressing Ganzfeld background of 1.61 log scotopic cd/m2. This indicates that dark-adapted WT responses to still dimmer stimuli are exclusively rod driven with minimal cone intrusion. Above this cone threshold intensity, the dark-adapted b-wave of WT will be a summation of rod and cone responses. Threshold versus intensity (TVI) studies gave no evidence of a rod influence on the mouse cone b-wave.