Evidence is presented for the existence of two forms of low-potential cytochrome a3. One appears to be similar to the low-spin form reported by Nicholls, P., and V. Hildebrandt (1978 Biochem. J. 173:65-72) and Wrigglesworth, J. M., J. Elsden, A. Chapman, N. Van der Water, and M. F. Grahn (1988. Biochim. Biophys. Acta. 936:452-464). It has a reduced Soret peak near 428 nm and a prominent alpha peak near 602 nm. This form is seen when the enzyme is either supplemented with lipoprotein or incorporated into a liposomal membrane, preexposed to a voltage greater than 400 mV for at least 30 min, and titrated in the presence of approximately 1 mM K3Fe(CN)6. The other form has a reduced Soret peak near 446 nm, and no prominent alpha peak. The 428-nm form has an Em near 175 mV and forms a CO complex with an Em near 225 mV. The 446-nm form has an Em near 200 mV and forms a CO complex with an Em near 335 mV.