The oxidation state (OS) formalism is a much-appreciated good in chemistry, receiving wide application. However, like all formalisms, limitations are inescapable, some of which have been recently explored. Providing a broader context, we discuss the OS and its interpretation from a computational perspective for transition metal (TM) complexes. We define a broadly applicable and easy-to-use procedure to derive OSs based on quantum chemical calculations, via the use of localized orbitals, dubbed the Intrinsic OS. Applying this approach to a cobalt complex in five OSs, isolated by Hunter and co-workers (Inorg. Chem. 2021, 60, 17445), we find that the calculated Intrinsic OS matches the formal OS, consistent with the experimental characterization. Through analysis of the delocalized orbitals, the ligand field of the Co(III) complex is found to be "inverted", despite every cobalt−ligand bond being classically dative from the localized perspective�a bonding scenario very similar to that of [Cu(CF 3 ) 4 ] − . This is not atypical but rather a natural consequence of these metals bonding in the high-valent region, and we propose a more restrictive definition of (locally) inverted bonding. Additionally, two bonding descriptors within the Intrinsic Bonding Orbital (IBO) framework (σ-gain and π-loss) are introduced, which enable facile quantification of electron-sharing covalency across a broad range of TM complexes.