The electron-transfer kinetics of cytochrome c oxidase were probed by measuring the reduction levels of bound cytochrome c, cytochrome a, and cytochrome a3 during steady-state turnover. Our experimental approach was to measure these reduction levels as a function of (1) the rate of electron input into tightly bound cytochrome c by varying the concentration of TMPD (N,N,N',N'-tetramethyl-p-phenylenediamine) and/or cytochrome c and (2) the rate of electron efflux out of cytochrome a (true Kcat) by changing the detergent surrounding cytochrome c oxidase. In most detergent environments, the rate of electron input into cytochrome c is not faster than the rate of electron efflux from cytochrome a. The relatively slow rate of electron input results in incomplete reduction of both cytochrome a and cytochrome c bound a the high-affinity site unless Kcat is very slow. When the high-affinity site is saturated with cytochrome c, the steady-state reduction level of cytochrome a defines Vmax,1, which is the maximum velocity of the high-affinity phase. The remaining fractional oxidation level of cytochrome a determines Vmax,2, the maximum velocity of the low-affinity phase. Therefore, it is the sum Vmax,1 + Vmax,2 which defines the maximum rate of electron transfer between cytochrome a and the bimetallic center, i.e., Kcat. We also were able to evaluate the true Kcat of cytochrome c oxidase in each detergent environment directly from the steady-state reduction levels without any of the complications introduced by the analysis of the polarographic kinetic data.(ABSTRACT TRUNCATED AT 250 WORDS)