Visual pigments are a class of proteins found in the membranes of photoreceptor cells (for reviews, see refs. 1 and 2). Their chromophoric unit is 11-cis-retinal covalently bound in the form of a Schiff base to the c-amino group of a lysine. The absorption of a photon by a visual pigment initiates a sequence of biochemical events that eventually lead to the generation of a neural signal by a photoreceptor cell. The identity of the primary photochemical event has been a subject of considerable interest and controversy. It was originally suggested that the primary event was an isomerization of the chromophore from its 1 1-cis to an all-trans conformation (3, 4). The strongest evidence favoring this mechanism was the observation (based on spectral data at low temperature) that an artificial pigment containing a 9-cis chromophore had the same photoproduct as rhodopsin itself. It was quite reasonably concluded that the most plausible common photoproduct formed from the two cis isomers is a trans isomer. A number of picosecond absorption studies of the primary event have raised widespread doubts as to the validity of the original model. It was found (5) that the primary event is complete in less than a few picoseconds at room temperature, and it was argued that this is too short a time for isomerization to occur [although other picosecond studies have reached the opposite conclusion (6)]. More recent evidence has come from the observation that at low temperature the rate of the process is significantly inhibited by deuterium replacement of the exchangeable protons on the pigment (7). Since only one proton on the chromophore is exchangeable, it is unlikely that this would have a measurable effect on the rate of isomerization. The picosecond measurements have generated numerous models (7-11) whose major feature is a photochemical proton transfer followed by a thermal cis-trans isomerization that occurs at a later stage. However, the original evidence upon which the suggestion of a cis-trans isomerization was originally based has never been discredited and, thus, it appears necessary to find a molecular model that is consistent with the entire body of available evidence. The purpose of this paper is to present such a model.There are, in fact, a fairly large number of observations that can be used in the construction and evaluation of alternative models. For example, we have shown, by using simple thermodynamic arguments, that a significant fraction of the photon's energy is "stored" in the primary photoproduct (12). Clearly a mechanism for energy storage must be an important component of any model that is proposed. Another energetic constraint may be derived from psychophysical and electrophysiological measurements of the level of thermal noise in photoreceptor cells. We show below that these observations may be interpreted directly in molecular terms and that they require an unusually high thermal activation energy for the primary event that appears to be inconsistent with many of the models that have been proposed.The ...