The Water Framework Directive (WFD, 2000/60/EC) mandates member states to consider trace element background values when establishing environmental quality standards (EQS) for sediments. This study proposes defining the "background" value as the trace element concentration that is consistently present in the unaltered natural environment, and the "reference" value as an upper limit of variation of naturally occurring concentrations, suggestive of anthropogenic enrichment if exceeded. We argue that background and reference values can be computed as the upper limits of the one-sided 95 % confidence interval for the median and for the 90th percentile, respectively. This proposition has been tested in two European river basins with different historical anthropogenic impacts: the Nalon (impacted by mining activities) and the Esva (mostly unpolluted) basins. For 110 bottom sediment samples collected from both pristine (background) and potentially impacted (non-background) locations, 19 elements were analyzed by ICP-MS/ICP-AES. The proposed background and reference values, calculated using the Bootstrap method, demonstrated consistency despite geological differences between the basins. The methodology proved particularly effective in datasets with a limited sample size and highly skewed data distributions. In the Nalon basin, with higher anthropogenic activity, the percentage of samples exceeding reference values was higher than in the Esva basin (e.g., 29 % vs 7 % for As and 11 % vs 0 % for Pb), suggesting the utility of these values in discerning potential impacts of historic mining activities. In the Esva basin, background and reference values were either near or below generic thresholds for good ecological status (LEL, TEL) unlike the Nalon basin, in which reference values for all elements, except three, exceeded these thresholds.