The photocycle of bacteriorhodopsin (BR) was studied at alkaline pH with a gated multichannel analyzer, in order to understand the origins of kinetic complexities in the rise and decay of the M intermediate. The results indicate that the biphasic rise and decay kinetics are unrelated to a photoreaction of the N intermediate of the BR photocycle, proposed earlier by others [Kouyama et al. (1988) Biochemistry 27, 5855-5863]. Rather, under conditions where N did not accumulate in appreciable amounts (high pH, low salt concentration), they were accounted for by conventional kinetic schemes. These contained reversible interconversions, either M in equilibrium with N in one of two parallel photocycles or L in equilibrium with as well as M in equilibrium with N in a single photocycle. Monomeric BR also showed these kinetic complications. Conditions were then created where N accumulated in a photo steady state (high pH, high salt concentration, background illumination). The apparent increase in the proportion of the slow M decay component by the background illumination could be quantitatively accounted for with the single photocycle model, by the mixing of the relaxation of the background light induced photo steady state with the inherent kinetics of the photocycle. Postulating a new M intermediate which is produced by the photoreaction of N was neither necessary nor warranted by the data. The difference spectra suggested instead that absorption of light by N generates only one intermediate, observable between 100 ns and 1 ms, which absorbs near 610 nm. Thus, the photoreaction of N resembles in some respects that of BR containing 13-cis-retinal.