Polyazines emerge as highly selective ligands toward actinide versus lanthanide separation. Electronic structures of several mono- and polyazine f-complexes of general formula MX3L (M+3 = Ce, Nd, Eu, U, Am, and Cm; X = RCp- or NO3-; L = N-donor ligand) related to Ln(III)/An(III) differentiation have been investigated using scalar relativistic ZORA/DFT calculations. In all cases, DFT calculations predict shorter An-N bonds than Ln-N ones whatever the azine used, in good agreement with available experimental data. The An-N bonds are also characterized by higher stretching frequencies than Ln-N bonds. The electronic structures of all species have been studied using different population analyses, among them natural population (NPA) and the quantum theory of atoms in molecule approach (QTAIM), as well as using different bond indices. The ability for Ln(III)/An(III) differentiation of the terdentate bipyrazolate BPPR ligand in the M(BPPR)(NO3)(3) complexes (M3+ = Ce, Eu, U and Am; R = H, 2,2-dimethylpropyl) where BPP = 2,6-bis(dialkyl-1H-pyrazol-3-yl)pyridine has been studied, with a special emphasis on the total metal-ligand bonding energy (TBE). The ZORA/DFT approach was found to properly reproduce the higher selectivity of the polyazine BPP ligand compared to monoazines, especially for the Eu-III/Am-III pair operating in spent nuclear fuel, using computed TBEs as criterion. Moreover, the orbital part of the total bonding energy appears also to rationalize well the observed selectivity.