An equation to estimate Hg concentrations of <4 μg/L in groundwaters of a polluted area in NE Italy was set out by using transplants of the aquatic moss Rhynchostegium riparioides as trace element bioaccumulators. The equation is derived from a previous mathematical model which was implemented under laboratory conditions. The work aimed at (1) checking the compliance of the uptake kinetics with the model, (2) improving/adapting the model for groundwater monitoring, (3) comparing the performances of two populations of moss collected from different sites, and (4) assessing the environmental impact of Hg contamination on a small river. The main factors affecting Hg uptake in the field were-as expected-water concentration and time of exposure, even though the uptake kinetics in the field were slightly different from those which were previously observed in the lab, since the redox environmental conditions influence the solubility of cationic Fe, which is a negative competitor of Hg(2+). The equation was improved by including the variable 'dissolved oxygen concentration'. A numerical parameter depending on the moss collection site was also provided, since the differences in uptake efficiency were observed between the two populations tested. Predicted Hg concentrations well fitted the values measured in situ (approximately ±50%), while a notable underestimation was observed when the equation was used to predict Hg concentration in a neighbouring river (-96%), probably due to the organic pollution which hampers metal uptake by mosses.