We have used suction-electrode recording to measure the early receptor current (ERC) from single, isolated mammalian photoreceptors. When a wild-type mouse rod was illuminated with light sufficient to close all the cGMP-gated channels, a succeeding bright laser flash bleaching a large proportion of the visual pigment produced an ERC, which at 37 degrees C consisted primarily of a single component of transient positive current. The amplitude of total charge movement of this component declined exponentially with successive flashes, consistent with the direct proportionality of the ERC to the quantity of pigment bleached. From the constant of exponential decline, it was possible to estimate the in vivo photosensitivity of mouse rhodopsin to be about 6 x 10(-9)microm(2) per molecule. We have also measured the ERC from rods of transducin-knockout mice, for which previous illumination to close the cGMP-gated channels was not required. The ERC of these rods was similar to that of wild-type rods but was followed by a slow component of outward current whose maximum amplitude in some cells approached that of the normal light response. This slow current was blocked by l-cis diltiazem, indicating that it was produced by ion flux through the cyclic nucleotide-gated channels of the outer segment; however, it could not have been produced by the normal transduction cascade, since it was recorded from rods lacking transducin. Since it was depressed by prior incorporation of the Ca(2+) buffer BAPTA, it was probably generated by light-activated Ca(2+) release earlier demonstrated in salamander and zebrafish. Recordings of the ERC from normal and mutant mice may provide a useful tool for the analysis of models of retinal disease, as well as exploration of the molecular origin of light-activated Ca(2+) release.