Electron transfer between Fe(CN)(6)(3-) and Fe(CN)(6)(4-) in homogeneous aqueous solution with K+ as the counterion normally proceeds almost exclusively by a K+-catalyzed pathway, but this can be suppressed, and the direct Fe(CN)(6)(3-)-Fe(CN)64-electron transfer path exposed, by complexing the K+ with crypt-2.2.2 or 18-crown-6. Fe((13)-CN)64- -NMR line broadening measurements using either crypt-2.2.2 or (with extrapolation to zero uncomplexed [K+]) 18-crown-6 gave consistent values for the rate constant and activation volume (k(0) (2.4 +/- 0.11) X 10(2) L mol(-1) s(-1) and DeltaV(0)(double dagger) = -11.3 +/-0.3 cm(3) mol(-1), respectively, at 25 degreesC and ionic strength 0.2 mol L-1) for the uncatalyzed electron transfer path. These values conform well to predictions based on Marcus theory. When [K+] was controlled with 18-crown-6, the observed rate constant k(ex) was a linear function of uncomplexed [K+], giving k(K) = (4.3 +/- 0.1) X 10(4) L-2 mol(-2) s(-1) at 25 degreesC and l = 0.26 mol L-1 for the K+-catalyzed pathway. When no complexing agent was present, k(ex) was roughly proportional to [K+](total), but the corresponding rate constant k(K)' (=k(ex)/[K+](total)) was about 60% larger than k(K), evidently because ion pairing by hydrated K+ lowered the anion-anion repulsions, Ionic strength as such had only a small effect on ko, kK, and kK'. The rate constants commonly cited in the literature for the Fe(CN)(6)(3-/4-) Self-exchange reaction are in fact k(K)'[K+](total) values for typical experimental [K+](total) levels.