The purpose of this study is to predict the spatio-temporal changes in land uses and to evaluate land-based pollutant loads in the future under Total Water Pollution Load Management System using CLUE-S model. For these ends, sensitive parameters of conversion elasticities in CLUE-S model were calibrated and these calibrated parameters of conversion elasticities, level II land cover map of year 2009, and 7 driving factors of land use changes were used in predicting future land uses in 2002 with two scenarios(Scenario 1: non area restriction, Scenario 2: area restriction). This projected land use map of 2020 was used to estimate land-based pollutant loads. It was expected that urban areas will increase in 2020 from both scenarios 1 and 2. In Scenario 1, urban areas are expected to increase within greenbelt areas and deforest would be expected. Under Scenario 2, these phenomena were not expected. Also the results of estimation of BOD and TP pollutant loads, the BOD difference between scenarios 1 and 2 was 719 kg/day in urban areas and TP difference was 17.60 kg/day in urban areas. As shown in this study, it was found that the CLUE-S model can be useful in future pollutant load estimations because of its capability of projecting future land uses considering various socio-economic driving factors and area-restriction factors, compared with conventionally used land use prediction model.