The experimental advances in electronic materials over the past decades have been accompanied by a remarkable increase in the ability to predict structural and electronic properties from first principles. Basic theory, along with modeling and simulation, has always been instrumental in understanding materials. Only recently, however, has the capability emerged to accurately predict properties based solely on the composition of the material, without any fitting to experimental quantities. Such a description must be based on a quantum-mechanical treatment, i.e., a solution of the Schrödinger equation for the system of atomic constituents. The seemingly impossible task of solving this vast many-body problem was rendered feasible by the development of density functional theory (DFT), an achievement for which Walter Kohn received the Nobel Prize in Chemistry in 1998. Other important developments that have greatly enhanced the ability to tackle large systems include pseudopotentials, the simultaneous optimization of electronic and atomic degrees of freedom as embodied in the Car-Parrinello method, and the tremendous increase in available computer power. In this talk I will focus on two areas in which these theoretical and computational advances have had a major impact, namely heterojunction interfaces and defects in semiconductors. Both are intimately connected to the high-quality growth techniques that have enabled a host of novel electronic devices. In the area of defects I will describe the effects of point defects and impurities on doping, specifically highlighting the role of hydrogen. A recently discovered universal alignment for the electronic level of hydrogen in semiconductors and insulators reveals a surprising link with the problem of heterojunction band lineups.