Experiments on local scouring around cylindrical piers were performed in steady clear-water conditions to investigate the effect of sediment size, D50, on equilibrium scour depth, dse. Thirteen experiments, run for 7 to 13 days, were carried out. The dependence of the normalized equilibrium scour depth, dse/Dp, on the relative sediment size, Dp/D50, was identified, corroborating the findings of Sheppard et al. (1995; 1999; 2004). Contrarily to consecrated literature and associated practice, it was observed that the normalized equilibrium scour depth decreases with the relative sediment size, for Dp/D50 > ~100. This effect is possibly one of the causes of frequently reported over prediction of scour by formulas established on the basis of laboratory studies, where frequently Dp/D50 < 100.