Mercury (Hg) contamination is a risk to environmental health, but predicting Hg exposure at the landscape-scale is difficult due to high variability in abiotic and biotic factors that influence Hg methylation and bioaccumulation. The Dragonfly Mercury Project (DMP) is a citizen-science program that monitors Hg at a continental scale in United States (U.S.) National Parks and other public lands using dragonfly larvae as biosentinels, providing significant spatiotemporal coverage of Hg accumulation data across a range of environmental conditions. Using Hg concentrations collected from more than 21,000 dragonflies (>1,200 site-years) across the U.S., we developed landscape-scale models to predict Hg bioaccumulation (i.e., exposure risk) using a Bayesian hierarchical modeling approach. The model variables include geospatial attributes (e.g., landcover, wetland extent, soil characteristics), water chemistry (e.g., DOC, pH, sulfate, nitrogen), and varying effects of habitat and ecoregion that account for context-dependent responses influencing Hg exposure. The goals for this modeling effort include predicting Hg risk in unsampled water bodies, forecasting risk by manipulating model inputs, iteratively updating the model with future collection events, and the integration of this model into a dashboard tool. We will present on the acquisition of ancillary data, constructing and training the model, model validation, and results.