A dynamic model to estimate the natural (non-anthropogenic) emissions of mercury from vegetation and water surfaces has been developed. The Mercury Re-Emissions Model (MREM) algorithms estimate the spatial and temporal distributions of mercury emissions over any specified spatial domain for various surface cover types. The emissions over land are a function of the land cover, transpiration rates, and temperature. The emissions over water are a function of the concentration gradient, the mixing of the air and water, and the temperature. This model will be used as a surface interface with the University of Connecticut regional air quality model for analysis of mercury transport, transformation and deposition.