The P+QA− and P+QB− charge recombination decay kinetics were studied in reaction centers from Rhodopseudomonas viridis reconstituted in phosphatidylcholine bilayer vesicles (proteoliposomes) and in chromatophores. P represents the primary electron donor, a dimer of bacteriochlorophyll; QA and QB are the primary and secondary stable quinone electron acceptors, respectively. In agreement with recent findings for reaction centers isolated in detergent [Sebban, P., & Wraight, C. A. (1989) Biochim. Biophys. Acta 9754–65] the P+QA− decay kinetics were biphasic (&fast and Kslow). Arrhenius plots of the kinetics were linear, in agreement with the hypothesis of a thermally activated process (probably via P+I−; I is the first electron acceptor, a bacteriopheophytin) for the P+QA− charge recombination. Similar activation free energies (∆G) for this process were found in chromatophores and in proteoliposomes. Significant pH dependences of kfast and kslow were observed in chromtophores and in proteoliposomes. In the pH range 5.5-11, the pH titration curves of &fast and kslow were interpreted in terms of the existence of three protonable groups, situated between I− and QA−, which modulate the free energy difference between P+I− and P+QA−. In proteoliposomes, a marked effect of o-phenanthroline was observed on two of the three pKs, shifting one of them by more than 2 pH units. On the basis of recent structural data, we suggest a possible interpretation for this effect, which is much smaller in Rhodobacter sphaeroides. The decay kinetics of P+Qb− were also biphasic. Marked pH dependences of the rate constants and of the relative proportions of both phases were also detected for these decays. The major conclusion of this work comes from the biphasicity of the P+QB−decay kinetics. We had suggested previously that biphasicity of the P+QA− charge recombination in Rps. viridis comes from nonequilibrium between protonation states of the reaction centers due to comparable rates of the protonation events and charge recombination. This hypothesis does not hold since the P+Qb−decays occur on a time scale ( ≈ 300 ms at pH 8) much longer than protonation events. This leads to the conclusion that kfast and A:slow (for both P+QA− and P+Qb−) are related to conformational states of the reaction centers, existing before the flash. In addition, the fast and slow decays of P+Qb− are related to those measured for P+QA−, via the calculations of the Qa−⇄Qb⇆ QaQb− apparent equilibrium constants, K2. Finally, these two “conformations”, which could arise from different interactions between Fe2+ (situated between QA and QB) and the quinone acceptors or between I and QA, behave as two independent components. They have their own AG, pH dependence, pH dependence of K2, and slightly different absorption change spectra, which can be separated near the isosbestic point at 833 nm. Ionic conditions, pH, and o-phenanthroline have notable effects on the relative proportions of the two phases. However, the meaning of these two populations is still unclear, as is their possible importance in the reaction center's function. © 1990, American Chemical Society. All rights reserved.