Most photovoltaic absorbers are identified using the standard Shockley-Queisser selection principle which relies on optimal band-gap values. However, this criterion has been shown to be insufficient, as many materials with appropriate values still perform badly. Here, we have employed calculations based on the density functional theory to assess three copper oxides as potential photovoltaic materials: Cu 2 O, Cu 4 O 3 , and CuO. Despite their promising theoretical solar power conversion efficiency of over 20%, experimental values are found to be far lower. Theoretical evaluation of the electronic and optical properties reveals that certain transitions within the band structures are dipole forbidden, whereas the fundamental band gaps of Cu 4 O 3 and CuO are of indirect nature. These findings correlate to the weak and shifted absorption properties found experimentally, which underpin the inefficient light capture by copper oxides. Based on these results and an applied extended selection metric, we can explain why copper oxides are unable to reach the efficiencies previously proposed theoretically and why we need to revise their maximum conversion values.