The single photon responses of retinal rod cells are remarkably reprod
ucible, allowing the number and timing of photon absorptions to be enc
oded accurately. This reproducibility is surprising because the elemen
tary response arises from a single rhodopsin molecule, and typically s
ignals from single molecules display large intertrial variations. We h
ave investigated the mechanisms that make the rod's elementary: respon
se reproducible. Our experiments indicate that reproducibility cannot
be explained by saturation within the transduction cascade, by Ca2+ fe
edback, or by feedback control of rhodopsin shutoff by any known eleme
nt of the cascade; We suggest instead that deactivation through a seri
es of previously unidentified transitions allows the catalytic:activit
y of a single rhodopsin molecule to decay with low variability. Two ob
servations are consistent with this view. First, the time course of rh
odopsin's catalytic activity could not be accounted for by the time re
quired for the known steps:in rhodopsin deactivation-phosphorylation a
nd arrestin binding; Second, the variability of the elementary respons
e increased When phosphorylation was made rate-limiting for rhodopsin
shutoff.