The electronic structures of the four- and five-coordinate aryl-substituted bis(imino)pyridine iron dinitrogen complexes, ((PDI)-P-iPr)FeN2 and ((PDI)-P-iPr)Fe(N-2)(2) ((TDI)-T-iPr = 2,6-(2,6-Pr-i(2)-C6H3-N=CMe)(2)C5H3N), have been investigated by a combination of spectroscopic techniques (NMR, Mossbauer, X-ray Absorption, and X-ray Emission) and DFT calculations. Homologation of the imine methyl backbone to ethyl or isopropyl groups resulted in the preparation of the new bis(imino)pyridine iron dinitrogen complexes, ((RPDI)-R-iPr)FeN2 ((RPDI)-R-iPr = 2,6-(2,6-Pr-i(2)-C6H3-N=CR)(2)C5H3N; R = Et, Pr-i), that are exclusively four coordinate both in the solid state and in solution. The spectroscopic and computational data establish that the ((RPDI)-R-iPr)FeN2 compounds are intermediate spin ferrous derivatives (S-Fe = 1) antiferromagnetically coupled to bis(imino)pyridine triplet diradical dianions (S-PDI = 1). While this ground state description is identical to that previously reported for ((PDI)-P-iPr)Fe(DMAP) (DMAP = 4-N,N-dimethylaminopyridine) and other four-coordinate iron compounds with principally sigma-donating ligands, the d-orbital energetics determine the degree of coupling of the metal-chelate magnetic orbitals resulting in different NMR spectroscopic behavior. For ((RPDI)-R-iPr)Fe(DMAP) and related compounds, this coupling is strong and results in temperature independent paramagnetism where a triplet excited state mixes with the singlet ground state via spin orbit coupling. In the ((RPDI)-R-iPr)FeN2 family, one of the iron singly occupied molecular orbitals (SOMOs) is essentially d(z)(2) in character resulting in poor overlap with the magnetic orbitals of the chelate, leading to thermal population of the triplet state and hence temperature dependent NMR behavior. The electronic structures of ((RPDI)-R-iPr)FeN2 and ((PDI)-P-iPr)Fe(DMAP) differ from ((PDI)-P-iPr)Fe(N-2)(2), a highly covalent molecule with a redox noninnocent chelate that is best described as a resonance hybrid between iron(0) and iron(II) canonical forms as originally proposed in 2004.