The impact of potentially toxic chemicals on wildlife is commonly assessed by comparing the intake of the contaminant with the no observable effects level (NOAEL) of intake. It is known, however, that there are considerable uncertainties inherent in this method. This study presents a Monte-Carlo based model to assess the degree of risk posed to birds (dunlin, Calidris alpina) from important estuarine habitats, and to show the limitations of such risk assessments, particularly with regard to data availability. The model was applied to predict the uptake of metals (Hg, Pb) in this shorebird species in Poole Harbour and the Severn Estuary/Bristol Channel, UK, two internationally important shorebird habitats. The results show that in both areas, Pb and Hg concentrations may pose an ecologically relevant toxic risk to wading birds. For Pb, uncertainty in NOAEL values dominates the overall uncertainty. Use of lethal toxicity data (LD50/100) was investigated as a method for assessing sub-lethal impacts from Hg. It was found that this method led to a significant under-estimate of the potential impact of Hg contamination, compared with direct estimation of NOAEL.