Data from Ka resonant inelastic X-ray scattering (RIXS) have been used to extract electronic structure information, i.e., the covalency of metal-ligand bonds, for four iron complexes using an experimentally based theoretical model. Ka RIXS involves resonant 1s -> 3d excitation and detection of the 2p -> 1s (K alpha) emission. This two-photon process reaches similar final states as single-photon L-edge (2p -> 3d) X-ray absorption spectroscopy (XAS), but involves only hard X-rays and can therefore be used to get high-resolution L-edge-like spectra for metal proteins, solution catalysts and their intermediates. To analyze the information content of Ka RIXS spectra, data have been collected for four characteristic sigma-donor and pi-back-donation complexes: ferrous tacn [Fe-11(tacn)(2)]Br-2, ferrocyanide [Fe-11(CN)(6)]K-4(4, ferric tacn [Felll(tacn)(2)Br-3 and ferricyanide [Fe-III(CN)(6)K-3. From these spectra metal ligand covalencies can be extracted using a charge-transfer multiplet model, without previous information from the L-edge XAS experiment. A direct comparison of L-edge XAS and Ka RIXS spectra show that the latter reaches additional final states, e.g., when exciting into the e(g) (sigma) orbitals, and the splitting between final states of different symmetry provides an extra dimension that makes Ka RIXS a more sensitive probe of a-bonding. Another key difference between L-edge XAS and Ka RIXS is the pi-back-bonding features in ferro- and ferricyanide that are significantly more intense in L-edge XAS compared to Ka RIXS. This shows that two methods are complementary in assigning electronic structure. The Ka RIXS approach can thus be used as a stand-alone method, in combination with L-edge XAS for strongly covalent systems that are difficult to probe by UV/vis spectroscopy, or as an extension to conventional absorption spectroscopy for a wide range of transition metal enzymes and catalysts.