1H NMR spectroscopy has been used to probe the various redox states of hemerythrin (Hr) to gain insight into the structural and magnetic changes that may occur in these states. In all, three oxidation states have been studied: deoxy [Fe(II),Fe(II)], semimet [Fe(II),Fe(III)], and met and oxy [Fe(III),Fe(III)]. The solvent-exchangeable imidazole NH protons of histidine ligands to the metal centers can be observed in all these complexes. For the met and oxy forms of the protein, these protons are found near 20 ppm, a shift significantly decreased from that found in mononuclear high-spin ferric imidazole complexes (ca. 100 ppm). This decrease in shift is consistent with the strong antiferromagnetic coupling (J = –100 cm–1) found for these complexes. For the deoxyHr complexes, the NH shifts range from 40 to 80 ppm, values comparable to those found for mononuclear high-spin ferrous complexes. This result shows the iron centers in the deoxy forms are not strongly coupled. Evans’ susceptibility measurements on deoxyHr and deoxyHrX (X = N3–, NCO−, F−) show the presence of a weak antiferromagnetic interaction (J = –15 cm−1) in deoxyHr and even weaker interactions in deoxyHrX (J = –11, –5, 0 cm−1 for X = F−, NCO−, N3−, respectively). For the semimetHr complexes, the NH shifts are also indicative of weak coupling. In all complexes except µ-SsemimetHr, J is estimated to be ca. –20 cm−1 from the temperature dependences of the shifts of the Fe(III)-coordinated imidazole NH’s. The NMR spectra also show that the iron atoms have a trapped valence formulation on the NMR time scale, with the anion coordinated to the ferric center. The small J values found for the deoxyHr and semimetHr complexes are consistent with the J values found for synthetic Fe(III)–Fe(III) and Fe(II)–Fe(II) complexes with µ-hydroxo-di-µ-acetato bridging units and suggest that the oxo bridge found in the met and oxy forms of hemerythrin has become protonated upon reduction. The persistence of a bridge between the metal centers in the various states of hemerythrin would facilitate the electron transfer required for reversible oxygenation. The NMR data support the mechanism proposed by Stenkamp et al. (Proc. Natl. Acad. Sci. U.S.A. 1985, 82, 713–716) for reversible oxygenation wherein the proton on the hydroxo bridge of deoxyHr is transferred to the bound peroxide as deoxyHr is converted to oxyHr. The stability of the semimetHrX complexes suggests that semimetHr superoxide [Fe(II),Fe(III) O2−] is a reasonable transition state in the oxygenation process.