It had been observed previously that a pair of transient EPR resonances (g = 1.78 and 1.69) appears within less than 5 ms on reoxidation of reduced cytochrome c oxidase by O2. Since the location of other lines that are part of the same signal was not known, the quantity of the paramagnetic species involved, and thus the significance of the observed resonances, remained questionable. We have now found a broad resonance at g = 5 which is obviously associated with those at g = 1.78 and 1.69. The width of the signal (approximately 250 mT) at the observed intensity suggests that it represents a significant fraction of one of the components of the enzyme. The signal disappears within less than 5 ms on addition of cyanide or sulfide but only within several hundred milliseconds after addition of ferrocytochrome c. This behavior suggests that it originates from the a3 component of the enzyme. It is suggested that the species represented in the signal is either identical with or part of what has been named collectively the "oxygenated" form and recently described "activated" forms of the enzyme. On reoxidation of reduced oxidase with oxygen enriched 90% in 17O, no change of signal shape was seen.