The kinetics of the primary electron transfer following excitation of the bacteriochlorophyll dimer (P) to its lowest excited singlet state were determined for a series of reaction center mutants of Rhodobacter sphaeroides that have P/P+ midpoint potentials 44-147 mV below wild type. These strains were capable of photosynthetic growth and had normal bacteriochlorophyll-to-bacteriopheophytin ratios. Decreasing the P/P+ midpoint potential resulted in an increase in the rate constant of initial electron transfer in each of the mutants tested. At room temperature, the fastest electron-transfer time constant observed was 1.8 ps from a mutant with a midpoint potential 127 mV below wild type. The dependence of the rate on driving force in these measurements and in previous measurements of mutants with high P/P+ midpoint potentials at room temperature was fit to a Marcus equation. Wild type was displaced approximately 100 meV from the peak of this curve. This analysis yielded a reorganization energy between 180 and 380 meV and an electronic coupling between 28 and 33 cm(-1) depending on what value is assumed for the standard reaction free energy of initial charge separation in wild-type reaction centers. However, the temperature dependence of both wild type and the high midpoint potential mutant reaction centers is much weaker than that expected from the activation energy predicted by the Marcus formalism. In fact, an activation energy of at least 15 meV is predicted for wild type which should completely prevent electron transfer at cryogenic temperatures, yet the rate constant of initial electron transfer is increased at 10 K. One explanation for this is that certain vibrational modes that promote electron transfer in the reaction center are coupled to light absorption and are not in thermal equilibrium with the surrounding bath on the time scale of electron transfer. Thus, part of the vibrational energy required for rapid initial electron transfer may come from the absorbed photon rather than from the surrounding bath.