The cyanide complex of Fe(II)Mb prepared and maintained at temperatures below 0 °C is sufficiently stable to permit spectroscopic characterization and allow comparison with free HCN and other ferric and ferrous CN complexes. The visible absorption spectrum of Fe(II)Mb-CN has a split α band maxima at 571 and 563 nm, suggesting distortion in the x-y plane of the porphyrin, Fe(II)Mb-CN, like the CO complex, was found to be optically active by circular dichroism. The C-N stretching frequencies for the CN-ferrous complexes are very sensitive to parameters within the heme pocket. The values are as follows: Fe(II)Mb at pH 8, 2057 cm-1 with a shoulder appearing at 2078 cm-1 at pH 5.6; Fe(II)Mp, 2034 cm-1. In contrast, the frequencies for C-N stretch differ little among ferric heme complexes, ranging from 2123 to 2125 cm-1 for myoglobin, hemoglobin, and microperoxidase. These values compare with free HCN (2094 cm-1) or CN- (2080 cm-1). Quantum chemical modeling of the neutral iron-porphyrin complex with imidazole and cyanide and of its anion was used to explain the effects of the cyanide coordination and of iron reduction on the C-N stretching frequencies. The lower vC-N for Fe(II)Mb-CN relative to the ferric complex is attributed to the appearance of additional electron density on all the anti-bonding CN orbitals. The extra electron density was also used to explain that the band width of C-N stretching mode was greater in the ferrous complexes than in the ferric complex. Finally, the calculation shows that o donation weakens the Fe-C bond, in qualitative agreement with the spontaneous dissociation of CN- from Fe(II)Mb at -5 °C. The sensitivity of CN complexes of ferrous heme proteins to the heme pocket environment and the ability to correlate spectroscopic parameters with calculated electron density suggest that infrared spectroscopy of the CN ligand is an appropriate tool to study ferrous heme proteins.