The Caspian Sea, a unique ecological system is under an increasing anthropogenic stress. Recently increased interest in development and exploitation of the Caspian Sea oil resources, along with the marked increase in the economic activities in both, agricultural and urban sectors within the entire Caspian Sea drainage basin brought to fore the potential increase in the anthropogenic environmental stress on the Caspian environmental system. The Caspian Sea is by far the largest lake in the world, with the coastline divided between five riparian countries: Azerbaijan, Iran, Kazakhstan, Russia and Turkmenistan. Its drainage basin stretches from sub-polar to sub-tropical climatic zones covering areas of eight states: Azerbaijan, Armenia, Georgia, Kazakhstan, Iran, Russia, Turkmenistan and Turkey. The principal objective of our project was to construct a model for predicting quantitative distribution of the non-conservative pollutants in time and space domains, by simulating all the transport and retardation processes of effluents introduced intentionally or inadvertently at any point in space and time into the Caspian Sea. These include petroleum products, pesticides and other chemicals introduced into the Sea by river outflow, atmospheric fallouts and dumping of untreated domestic and industrial wastes. The model accounts for all hydro-mechanical and hydro-chemical processes affecting temporal concentration of non-conservative compounds, and is based on two extensive data bases: (1) seasonal and spatial distribution of the physical-chemical and hydro-meteorological parameters within the Caspian Sea Basin; and (2) physical-chemical parameters of various hydrocarbons and other pollutants. The transport module of the model takes predetermined current and turbulent diffusivities and uses Langrangian tracking to predict the motion of the individual particles (droplets), the sum of which constitute the simulated spill. Currents and turbulent diffusivities used in the simulations are generated by a numerical ocean circulation model (Princeton Ocean Model) implemented for the Caspian Sea. The model may be easily adapted for use in the other great lakes of the world. Acknowledgement: this project was supported by CRDF Grant#2284