The electronic structures and magnetic hyperfine interactions in a series of related heme compounds have been studied. Four of these are high spin ferric heme compounds with different fifth ligands, namely, fluoride, chloride, bromide, and hydroxide, and the fifith is a low spin cyano–ferric–heme compound. The molecular orbitals needed in the calculation were obtained by the self-consistent charge extended Hückel method. For the four high spin ferric compounds the hyperfine constants of 57Fe and 14N nuclei were found to vary only 4% among themselves. In the chloride derivative, the only one for which 57Fe data are available, excellent agreement was found between our result and the 57Fe hyperfine field in hemin from Mössbauer measurements. For 14N nuclei in the fluoride, chloride, and bromide derivatives the near constancy of the predicted hyperfine constants is borne out experimentally, with the theoretical values being all about 50% of experiment. For the low-spin cyano compound we predict the 57Fe hyperfine field per spin to be 60% larger than the corresponding fields in high spin heme derivatives, while the hyperfine field at 14N per spin is substantially smaller than its counterpart in the high spin compounds. Our results for the hyperfine fields in the series of compounds studied and their agreement with available experimental data demonstrate that there is substantial delocalization of unpaired spin away from iron for these ferric heme compounds, especially the high spin ones. This feature is also supported by a similar calculated decrease in the iron 3s core energy splittings relative to that expected in free Fe+3 ion, which is verified by ESCA measurements in hemin chloride. The need for additional experimental data on the hyperfine fields, particularly for 57Fe and 14N in the cyano compound, is stressed.