The first discernible intermediate when fully reduced cytochrome c oxidase reacts with O-2 is a dioxygen adduct (compound A) of the binuclear heme iron-copper center. The subsequent decay of compound A is associated with transfer of an electron from the low-spin heme a to this center. This reaction eventually produces the ferryl state (F) of this center, but whether an intermediate state may be observed between A and F has been the subject of some controversy. Here we show, using both optical and EPR spectroscopy, that such an intermediate (P-R) indeed exists and that it exhibits spectroscopic properties quite distinct from F. The optical spectrum of P-R is Similar or identical to the spectrum of the P-M intermediate that is formed after compound A when two-electron-reduced enzyme reacts with O-2. An unusual EPR spectrum with features of a Cu-B(II) ion that interacts magnetically with a nearby paramagnet [cf. Hansson, O., Karlsson, B., Aasa, R., Vanngard, T., and Malmstrom, B.G (1982) EMBO J. 1, 1295-1297; Blair, D. F., Witt, S. N., and Chan, S. I. (1985) J. Am. Chem. Sec. 107, 7389-7399] can be uniquely assigned to the P-R intermediate, not being found in either the P-M or F intermediate. The binuclear center in the P-R State may be assigned as having an Fe-a3(IV)=O Cu-B(II) structure, as in both the P-M and F states. The spectroscopic differences between these three intermediates are evaluated. The P-R State has a key role as an initiator of proton translocation by the enzyme, and the thermodynamic and electrostatic bases for this are discussed.