Perovskite structured oxides are highly versatile in its application, especially as electrodes in heterostructures: for example in the development of solid oxide fuel cells (SOFC), which are promising candidates for power plants and climate neutral ways of locomotion. With specific doping, the properties of perovskite oxides can be changed such that e.g. a metal-insulator transition occurs. The description of this behavior at the example of doped CaRuO3 is the main aim of this thesis. A very important point in the description of doped crystal structures is the randomness of the impurity atoms. P. W. Anderson suggested that the random distribution of impurities within a tight-binding model leads to a localization of the wavefunction. Numerical results of the Anderson model and its application to real systems are given. The calculation of electronic properties of perovskites is done with methods of density functional theory (DFT). The usage of the projector augmented wave formalism implemented in VASP makes it possible to predict tilting and distortion of the perovskite structure by ionic relaxation. The effects of Co-doping are described with supercell methods (VASP) or the coherent potential approximation (CPA) implemented in a Korringa, Kohn and Rostoker Green's function method. The CPA allows the calculation of electronic properties of arbitrary impurity concentrations in a perturbative regime. The model approach (Anderson tight-binding) and the DFT calculations show indeed strong indications for a metal-insulator transition with increasing substitutional disorder. The electrical conductivity is suppressed because of localization at the impurity sites. This can also be observed by energy dispersion and density of states calculations. Flat bands at the Fermi level indicate a wave function localization and hence a drop in electrical conductivity. Those results agree with recent experimental measurements.