The design of more efficient organic photovoltaics starts with an increase in understanding of the fundamental processes related to organic photovoltaics, such as the charge separation processes at the organic/organic interface, which can only be remedied by a combined theoretical and experimental effort. In this thesis we use a variety of computational techniques to address current questions in the field or organic photovoltaics. Applying the [delta]SCF method to a test set of conjugated organic molecules we find it has an error of +/-0.3 eV, and by using the [delta]SCF wavefunctions for a multi-reference basis we construct a new perturb then diagonalize multi-reference perturbation theory method that performs well for both ground and excited state potential energy surfaces, called [delta]SCF(2). Our computed singlet fission rates are in near quantitative agreement with experimental measurements in a variety of pentacene derivatives, and we find that the singlet fission mechanism proceeds through a non-adiabatic to adiabatic transition. By combining ab initio rate constants and Kinetic Monti-Carlo we get an accurate prediction of triplet diffusion and show that only a small decrease occurs when the crystal becomes highly disordered, and no significant traps exist. Our models of the organic/organic interface reveals that the the simple picture of constant HOMO and LUMO levels throughout an organic photovoltaic device is only qualitatively accurate at best. At the organic/organic interface effects such as change in the dielectric constant, decreased packing efficiency, and molecular multipole moments all can contribute to changing the HOMO and LUMO levels at the interface by over 0.2 eV, which is large enough to drive apart thermally relaxed charge transfer states at the interface. The work in this thesis provides insight into how to achieve better exciton diffusion and charge separation in organic photovoltaics, as well as insight into a number of electronic processes relevant to organic photovoltaics.