Magnetic resonance spectroscopy comprises some of the most powerful analytic tools in materials science available to date. The ongoing development of this experimental technique will significantly widen the scope of the method and the information that can be extracted. In this project, we want to enhance the accuracy of calculated EPR parameters, necessary for a complete analysis of the experimental spectra. The systems to be investigated will range from silicon-related surfaces and interfaces, as e.g. realized in solar cells, to metal-organic macrocycles like porphyrins adsorbed on various substrates.The geometry and electronic structure of the molecules are strongly influenced by the substrate. Hence, the substrate has to be included explicitely into the theoretical modelling, e.g. by periodic boundary conditions. This necessitates serious methodological extensions. In tight collaboration between theory and experiment, our new concepts will lead to a more reliable interpretation of the experimental data. The resulting better understanding of the physics behind the magnetic fingerprint of the spin systems will lead to optimized sample preparation and, by this, to increased experimental sensitivity.