Abstract The optical constants of many metals commonly used in solar cells, e.g. as contacts, rear side planar reflectors, or more complex nanopatterned light-trapping structures, can vary depending on deposition method, thickness and other factors, and as such are not documented consistently in the literature. In the case of nanometallic light-trapping structures specifically designed to improve absorption in a solar cell, the choice of optical constants used in simulations significantly affects the predicted enhancement, as well as the structure's optimal dimensions. The trade-off between coupling into guided modes in the photovoltaic material and the number of photons absorbed parasitically in the metal leads to small differences in the optical constants giving significantly different results for the quantum efficiency and photogenerated current. This work documents several optical constant sources for silver, aluminium, gold and titanium, and the effect this has on plasmon quality factors. The effect of choosing different optical constant sources on modelling outcomes is quantified by considering the optimization of a test structure comprising a grid of metal nanodisks on the front surface of a thinned-down GaAs cell. Finally, we define a new spectrally-integrated figure of merit for comparing the expected performance of metals in light-trapping structures based on their optical constants, which we name the spectral absorption enhancement factor (SAEF).