The spectral properties of a P-450 preparation from adrenocortical mitochondria were compared with those of both other hemoproteins and model compounds to examine the possibility suggested by previous authors that the mercaptide anion (thiolate anion, RS-) of a cysteine residue might be one of the axial ligands of P-45O. 1. The following features were found in addition to the well known anomalous spectral characteristics of P-450. The Soret band of the ferrous form was at anomalously short wavelengths compared with those of other ferrous high-spin hemoproteins. The regularity in the shifts of the Soret maximum observed with various forms of usual hemoproteins was disrupted in the case of P-450 by the maximum of the ferrous form. 2. The complex of methemoglobin with 2-mercaptoethanol, which was the model for P-450 used by previous authors for spectral and ESR studies, was studied more precisely. The model showed an absorption spectrum very similar to that of the imidazole complex of ferric P-450, whereas the complex of methemoglobin with imidazole showed both Soret and visible absorption maxima at shorter wavelengths. The rapid rate of reduction of methemoglobin by dithionite decreased on binding to mercaptoethanol and the slow rate of reduction of P-450 increased on binding to imidazole. Thus the half-reduction times for both complexes were of the same order of magnitude. 3. An absorption spectrum which was very similar to the spectrum of ferric high-spin P-45O could be obtained by quick scanning immediately after ferric protoheme had been dissolved in a 33% solution of mercaptoethanol in water containing Tween 20 and NaOH. 4. Protoheme dissolved in an aqueous solution of Nikkol BL9EX (a non-ionic detergent) containing NaOH, dithionite, and 1 mM mercaptoethanol formed a transient intermediate complex having a Soret maximum at 470 nm. Treatment of this complex with carbon monoxide resulted in the formation of a CO adduct having a Soret maximum at 450 nm and a single band at about 555 nm in the visible region. These maxima were similar to those of the carbon monoxide complex of P-450. Many of the results could be explained readily by assuming that a mercaptide anion was one of the axial ligands, and also that heme was held in a hydrophobic crevice in the P-450 molecule. The coordination structure of P-450 seems to be more flexible than those of usual hemoproteins.