The record laboratory cell (∼1 cm2 area) efficiency for thin-film cadmium telluride (CdTe) is 16.5%, and that for a copper indium diselenide (CuInSe2) thin-film alloy is 19.5%. Commercially produced CdTe and CuInSe2 modules (0.5–1 m2 area) have efficiencies in the 7–11% range. Research is needed both to increase laboratory cell efficiencies and to bring those small - area efficiencies to large-area production. Increases in laboratory CdTe cell efficiency will require increasing open-circuit voltage, which will allow cells to harvest more energy from each absorbed photon. This will require extending the minority carrier lifetime from its present τ ≤ 2 ns to τ ≥ 10 ns and increasing hole concentration in the CdTe beyond 1015 cm2, which appears to be limited by compensating defects. Increasing laboratory CuInSe2-based cell efficiency significantly beyond 19.5% will also require increasing the open-circuit voltage, either by increasing the bandgap, the doping level, or the minority carrier lifetime. The photovoltaic cells in commercial modules occupy tens of square centimeters, and both models and experiments have shown that low-performing regions in small fractions of a cell can significantly reduce the overall cell per formance. Increases in commercial module efficiency will require control of materials properties across large deposition areas in a high-throughput environment to minimize such non-uniformities. This article discusses approaches used and research needed to increase the ultimate efficiencies of CdTe- and CuInSe2-based devices and translate these gains to commercial photovoltaic modules.