We explore the interrelationship between (i) the kinetic data for the primary electron transfer (ET) from the excited singlet state (1P*) of the bacteriochlorophyll dimer (P) to bacteriopheophytin (H) in the reaction centers (RCs) of purple photosynthetic bacteria and (ii) the energetics and dynamics of the primary radical pair, P+H, i.e., its singlet‐triplet splitting rate, J, and triplet recombination rate, kT. We present a critical scrutiny of the ET mechanisms with regard to the role of the accessory bacteriochlorophyll (B). On the basis of the very weak temperature dependence of J, which sets a lower limit to the energy of (hypothetic) transient short‐lived ionic intermediate states, we argue against the applicability of sequential ET mechanisms involving P+B− H or P+BH−. The nonadiabatic/adiabatic ET mechanism is fraught with difficulties pertaining to a very low medium reorganization energy for the nonadiabatic crossing and an exceedingly high rate of the adiabatic crossing, together with a predicted order‐of‐magnitude enhancement of the low‐temperature (77 K) fluorescence quantum yield in high electric fields (5–10 mV/Å) which seem not to be borne out by experiment We propose that primary ET occurs via the unistep superexchange 1P*BH r̊ P+H−, which is mediated by electronic coupling with P+B−H. The analysis of the competition between superexchange and thermally activated ET resulted in large values of the intermolecular electronic couplings at the equilibrium nuclear configuration of 1P*BH. We suggest that the structural relaxation of one of the prosthetic groups (H) accompanying the primary ET results in the decrease of the intermolecular electronic coupling at the equilibrium configuration of P+BH−. It was established that the superexchange‐predicted kinetic data are consistent with the magnetic data for the radical pair. The small value of J for P+BH− originates from the cumulative effect of: (i) an essential, model‐independent cancellation between the triplet contribution, which is related to kT by a self‐consistency relationship, and the singlet contribution; (ii) the effect of configurational relaxation on the singlet contribution. Our analysis supports the superexchange mechanism for primary ET in bacterial photosynthetic RCs.