Some remarks are made on the validity of a commonly used analytical model based on the rigorous coupled wave analysis to describe the optical response of one-dimensional metallic gratings.