AbstractLead halide perovskites have prompted great interest, offering impressive photovoltaic performances. Most fundamental investigations and cell optimizations focus on solution‐based solar cells, which are not easily extended to larger scales. Commonly in these cells, losses in the open‐circuit voltage are attributed to arise primarily from interface recombination, and therefore the most studies have focused on optimization of the surface to eliminate defects states. In contrast, thermal evaporation is an alternative, solvent‐free, and scalable method to deposit lead halide perovskites that is gaining attention. However, the number of reports showing high‐efficiency solar cells (> 20%) prepared using thermal evaporation is still small. Here, the origins of non‐radiative charge carrier recombination are investigated in perovskite cells that are deposited via thermal co‐evaporation. This is done through a combination of photoluminescence spectroscopy, current‐voltage characterization, and simulations. It is found that the non‐radiative recombination in these cells is caused equally by bulk and interface defects. In general, it is advocated to perform a dual analysis of the photoluminescence spectroscopy of both the film and the photovoltaic device, in conjunction with current‐voltage measurements. It is emphasized that such a dual analysis is needed to enable the identification of improvements and to unlock further advancements in this technology.