Light responses were recorded from toad rods in order to study the dependence of the kinetics and amplitude of the single photon response on the longitudinal position of excitation within the rod outer segment. Membrane current was recorded from the inner segment of an isolated rod with a suction electrode while stimulating the outer segment with a dim transverse slit of light. Flashes at the tip of the outer segment gave smaller average responses than flashes at the base. Comparison of amplitude histograms from the two positions revealed that the fraction of incident photons eliciting an electrical response was the same at tip and base. Characteristic differences in flash sensitivity are therefore attributed to differences in the amplitude of the single photon response. Flash responses from the tip were slower than those from the base. For most cells, the tip response could be fitted by the same multistage filter equation that fitted the base response when only one of the filter time constants was increased. For both tip and base responses, the ensemble variance as a function of time was proportional to the square of the ensemble average. This indicates that single photon responses had the same wave form as their respective averages, and that there was no significant contribution of fluctuations in response latency to the wave form of the average. Background light reduced flash sensitivity at the tip more than at the base. The calculated cable attenuation of rod outer segments is not sufficient to explain the observed differences between the responses of the tip and base. The differences might instead be associated with a longitudinal gradient of internal sodium concentration, or with ageing of the outer segment discs. Calculations suggest that in the intact eye, the amplitude and time course of the average electrical response to absorption of a photon should depend slightly on the wave-length of the photon.