In native RCs from Rb. sphaeroides the recombination D(+)Q(A)Q(B) --> DQ(A)Q(B) proceeds via an indirect path involving the intermediate state D(+)Q(A)Q(B). To observe the direct recombination rate, k(BD), the energy difference between the D(+)Q(A)Q(B) and D(+)Q(A)Q(B)(-) states has to be increased. This had been accomplished in mutant RCs (DN(L213)) by lowering the energy of the D(+)Q(A)Q(B)(-) state [A. Labahn, M.L. Paddock, P.H. McPherson, M.Y. Okamura and G. Feher, J. Phys. Chem. 98 (1994) 3417] or, as presented in this work, by raising the energy of the D(+)Q(A)(-)Q(B) state through substitution of Q(10) by the low potential quinones: (2,3,5-trimethyl-1,4-naphthoquinone, 2,3,6,7-tetramethyl-1,4-naphthoquinone, 2-chloro-9,10-anthraquinone) while retaining the native Q(10) in the Q(B) site. The recombination rates k(BD) in these hybrid RCs were fitted with the Marcus theory giving a reorganization energy, lambda(BD) = 1.1 +/- 0.1 eV and an electronic matrix element V(r) = (1.2 +/- 0.5) X 10(-8) eV. The larger value of lambda(BD) compared to lambda(AD) (1.1 versus 0.6 eV) is consistent with the more polar environment of Q(B)(-) and is believed to be the main contributor to the large observed ratio of k(AD)/k(BD) approximate to 100.