Total gaseous mercury (TGM) air-water flux measurements were taken using a dynamic flux chamber (DFC) coupled with a gaseous mercury (Hg) analyzer at the Bay St. Fran¿ois (BSF) wetlands (Quebec, Canada) in summer 2003. The measured TGM fluxes over water exhibited a consistent diurnal pattern, with maximum emissions during daytime and minimum fluxes occurring at night. Pearson correlation analysis showed that solar radiation was the most influential environmental parameter in TGM air-water exchange. Significant correlations were also found between TGM fluxes and 1 hour time-lagged water temperature, indicating the enhancement of fluxes by bacterial activities or chemical reactions. The concentrations of dissolved gaseous mercury (DGM) in water were measured during the 2003 sampling period and indicated that DGM was always supersaturated, which implied that the water body acted primarily as a source of mercury to the atmosphere. Several empirical models of mercury air-water gas exchange were developed and evaluated. Compared to the published models, these proposed models were capable of producing good results, leading to a better agreement between the measured and modeled fluxes (improvements by 48--98%). Among these empirical models, the ones linking TGM fluxes with net radiation were superior because of their strong predictive capability. Two preferred models were selected for air-water TGM flux estimation from Lake St. Pierre's surrounding wetlands. These two models yield a mean emission of 0.19--0.24 kg mercury during May--September each year from 1999 to 2003.