This paper quantifies atmospheric mercury (Hg) emissions from substrates and fumaroles associated with three hydrothermal systems: Lassen Volcanic Center, California (LVC); Yellowstone Caldera, Wyoming (YC); and Dixie Valley, Nevada (DV). Substrate Hg fluxes were measured using field chamber methods at thermal and nonthermal sites. The highest Hg fluxes (up to 541 ng m⁻² h⁻¹) were measured at thermal active areas. Fluxes from altered and unaltered nonthermal sites were <10 ng m⁻² h⁻¹ and were comparable to those measured in natural low-Hg background regions for YC and DV, and at LVC they were slightly elevated. Similarly, reactive gaseous mercury concentrations were higher in thermal active areas. Using a Geographic Information System framework for scaling, estimated area-average Hg emissions from substrates were 0.9–3.8 ng m⁻² h⁻¹ at LVC, 0.8–2.8 ng m⁻² h⁻¹ at YC, and 0.4–0.5 ng m⁻² h⁻¹ at DV. At LVC, nonthermal areas were the primary substrate source of atmospheric Hg (>98%). At YC, substrate Hg emissions were dominated (50 to 90%) by acidically altered thermal areas. Substrate emissions at DV were low and primarily from nonthermal areas (66% to 75%). Fumarole emissions at LVC (91–146 kg yr⁻¹) and YC (0.18–1.6 kg yr⁻¹ for Mud Volcano) were estimated by applying Hg:H₂O and Hg:CO₂ ratios in hydrothermal gas samples to H₂O and CO₂ emissions. Applying total area-average emissions from substrates and thermal features at LVC, YC, and DV to similar systems across the conterminous United States, yearly atmospheric Hg emissions from active hydrothermal systems are projected to be 1.3–2.1 Mg.