We propose an unexplored class of absorbing materials for high-efficiency solar cells: heterostructures of transition-metal oxides. In particular, LaVO(3) grown on SrTiO(3) has a direct band gap ∼1.1 eV in the optimal range as well as an internal potential gradient, which can greatly help to separate the photogenerated electron-hole pairs. Furthermore, oxide heterostructures afford the flexibility to combine LaVO(3) with other materials such as LaFeO(3) in order to achieve even higher efficiencies with band-gap graded solar cells. We use density-functional theory to demonstrate these features.