The nutrient concentration of river water was obtained in the Tokachi River tributaries (Toshibetsu, Satsunai, Otofuke, Shikaribetsu and Sarubetsu) in 2004 , 2005 and 2005 . Based on the discharge, precipitation and other meteorological data (air temperature, humidity, atmospheric pressure, wind and radiation), a tank model was applied to analyze runoffs of each river. In order to understand the mechanism of nutrient supply from the catchment area, we coupled the land use patterns, soil types and geological distributions with the discharge patterns by the model. From the tank model, the total discharge in winter consisted of more than 80% baseflow. In the snowmelt season, when the discharge increased drastically, the contribution of surface runoff increased up to more than 50% of the total discharge. The contribution of the surface runoff was highly correlated with the effective rainfall in summer, indicating that the rainfall event is immediately reflected by the increase in river discharge. In the Toshibetsu River, the silicic acid concentration varied seasonally with the large amplitude, and was negatively correlated with the discharge. The concentration tended to be high in catchments of the volcanic crust cover. During the winter low discharge, all the rivers showed the highest silicic acid concentration. However, during a spring thaw, the concentration decreased consistently. This suggests that the silicic acid concentration is determined by the mixing between the baseflow of highest concentration and the surface runoff of relatively low concentration. The nitrate concentration differed largely among rivers, which ranged from high concentration ( 300- 400 μmol l^-1) in the Sarubetsu River to low concentration (20- 50 μmol l^-1) in the Toshibetsu River. The concentration was positively correlated with the proportion of the farming area in catchments. This suggests that unutilized excess fertilizer is the major source of nitrate to rivers. The nutrients concentration variations in the Tokachi River tributaries are judged to be affected strongly by the land use, soil and geology rather than discharge patterns.