We use a multimedia mass balance model to assess mercury cycling in a contaminated tropical montane forest section, in Kodaikanal in South India. The forest section received more than 1,300 kg of direct mercury release from a now-defunct thermometer manufacturing factory between 1984 and 2002, and continues to receive rainfall runoff from the factory. Modeling results suggest that a complete or partial cleanup of factory soil in 2020 will marginally reduce the level of contamination of forest soil by 2050 compared with a no-cleanup scenario. Mass balance analysis showed that the dominant source of mercury to forest changed from atmospheric deposition in the uncontaminated forest to direct inputs and runoff of mercury from the factory. Monte Carlo uncertainty analysis showed that the most sensitive parameters influencing the fate of mercury in the forest (for example, mass in soil or vegetation, and runoff to downstream) changed from those describing atmospheric deposition in the uncontaminated forest to those describing direct inputs and runoff of mercury from factory or forest soil and the partitioning of mercury to soil. Forest soil now acts and will continue to act as a source of mercury to downstream locations. For a better assessment of the effects of mercury pollution from any future contaminated sites, parameters such as sediment runoff, partitioning of mercury to soil, concentration of mercury species in air, and concentration of mercury in factory and forest soil must be regularly measured. © 2020 American Society of Civil Engineers.