Seven Wisconsin rivers with contrasting, relatively homogeneous watershed composition were selected to assess the factors controlling mercury transport. Together, these watersheds allow comparisons of wetland, forest, urban and agricultural land-uses. Each site was sampled nine times between September 1993 and September 1994 to establish seasonal signatures and transport processes of total mercury (Hg-T) and methylmercury (MeHg). Our results clearly show that land use and land cover strongly influence mercury transport processes. Under base-flow conditions, unfiltered MeHg yield varies by a factor of sixteen (12-195 mg km(-2) d(-1)), and increases with the fraction of wetland area in the watershed. Elevated mercury yields during high flow are particle-phase associated in agricultural sites, but filtered-phase associated in wetland sites. Methylmercury represented less than 5% of total mercury mobilized during the spring thaw across all watersheds. Autumn MeHg yield was generally 11-15% of Hg-T in wetland influenced watersheds, though a maximum of 51% was observed. In some cases, single high-flow events may dominate the annual export of mercury from a watershed. For example, one high-flow event on the agricultural Rattlesnake Creek had the largest Hg-T and MeHg yield in the study (107 and 2.32 mg km(-2) d(-1), respectively). The mass of mercury transported downstream by this single event was an order of magnitude larger than the eight other (non-event) sampling dates combined. These results underscore the importance of watershed characteristics and seasonal events on the fate of mercury in fresh water rivers.