Since the introduction of the Angular overlap model (AOM) in the mid-1960s, expressing d orbital energies in terms of the s-and p-antibonding parameters e s and e p , the AOM has failed to supplant crystal field theory as the standard model to explain structure and electronic spectra in transition metal complexes. This is so despite the much more obvious connection in the AOM between structure and d orbital energies, the pictorial simplicity of the AOM approach, and the more consistent transferability of AOM parameters from one complex to another. The main reason is probably that AOM parameters cannot be determined uniquely when all the ligands are on the Cartesian axes. The scales for e s and e p must then be fixed arbitrarily, as is done automatically in the crystal field model. A number of experimental approaches have evolved to solve, or at least evade, the nonuniqueness problem, including: (a) the assignment of e p for saturated amines to zero, reflecting their inability to p-bond; (b) the simultaneous use of magnetic and spectroscopic data; (c) the inclusion of data from sharp, spinforbidden lines in Cr(III) spectra, along with application of the exact geometry and full d n configuration interaction in computations, or any combination of these; (d) the use of charge transfer bands involving dp orbitals to determine e p values. As of yet, the level of consistency among different techniques leaves something to be desired, and even with a common technique, reported AOM parameter values for particular metal-ligand combinations show a much higher variability than one would like, given even minimum expectations for transferability. Some of the variability can be ascribed to differences in the other ligands present. Even with these variations, AOM parameter sets can be usefully correlated with kinetic and thermodynamic data from both photochemical and thermal reactions.