This paper highlights the role played by surface roughness on the resuspension of nano- and micro-sized particles and, in particular, the need to extract more information from measurements of the surface profile than typical values such as the average roughness R-a and the rms roughness R-rms (usually obtained through AFM or SEM measures). For that purpose, standard experimental measurements of surface roughness are analysed. Then, numerical results obtained with a stochastic model for particle resuspension are analysed and compared to experimental data. This analysis reveals that particle resuspension can only be properly captured with more detailed representations of surface roughness that include information on the distribution of the curvature radius and surface coverage of roughness features.