This paper analyzes the impact of land use factors to pollutant loadin the watershed of Lake Kasumigaura. In the field experiment, two types of optical signals and water quantity were used to estimate the loads of chemical oxygen demand (COD), particulate chemical oxygen demand (P-COD), total nitrogen (TN), nitrate (NO3-N), total phosphorus (TP) and particulate phosphorus (P-P). Annual pollutant loads were calculated by the Artificial Neural Network (ANN) model simulation results. Regression analysis was applied to evaluate the impact of different types of land use in different periods to the 6 pollutants respectively. Dependence of pollutant loads of land use characteristics was found in the tributary basins of the Kasumigaura.