AbstractA microcosm consisting of water, sediment, and air compartments was used to simulate mercury geochemical cycling in a mercury‐contaminated (μg L−1) pond at Oak Ridge, Tennessee. Total and dissolved mercury and total methylmercury were analyzed in water and sediment; total gaseous mercury and head‐space mercury were analyzed in the water and head space, respectively. The production of gaseous mercury was correlated to dissolved mercury (0.2‐μm filtration), and methylmercury was mainly produced in the sediment compartment. Addition of mercuric chloride to the system increased the production of head‐space mercury by a factor of 10 but did not affect the methylation rate. Saturation of gaseous mercury in microcosm water varied from 480 to 1,500% of the solubility of elemental mercury and was controlled by unidentified factors. The microcosm maintained stable conditions for up to 3 weeks, and a mass balance indicated that it reasonably simulated the cycling of mercury in the pond. This microcosm could be used to test remedial treatments aimed at decreasing the amount of mercury that is available for accumulati