Changes of fundamental climate and hydrological cycle quantities in the Upper Mississippi River Basin (UMRB) from the 20th to the 21st century are studied with the Soil and Water Assessment Tool (SWAT) driven by daily meteorological data from contemporary climate (20C) and future scenario (A2) simulations of 10 global climate models in the IPCC Data Archive and NARCCAP regional climate models. GCM ensemble results show that with the warming from the 20th to the 21st century in the UMRB in all seasons, precipitation decreases in summer but increases in all other seasons. ET has a lag of a month in change, and only decreases in July-September. Snowmelt decreases in spring, possibly due to the lower fraction of snow in total precipitation. Streamflow increases in winter but decreases in all other seasons. Correlation analysis for interannual variations in both contemporary and future climates indicates that summer precipitation has a significant negative correlation with temperature, meaning that higher summer temperature makes saturation more difficult. Streamflow has positive correlation with precipitation in all seasons, and its correlation with temperature is positive in winter but negative in all other seasons. The year-to-year change of ET depends strongly on precipitation in August-October but on temperature in all other months. Snowmelt has positive correlation with temperature in winter but negative in both spring and fall, and its correlation with precipitation is always positive. The magnitudes of the change from contemporary to future climates are compared with the interannual climate variability, and the above interannual relationships are used to understand how the long-term change of hydrological cycle is affected by the warming climate. It is revealed that, compared with interannual variations, the long-term changes of streamflow and other hydrological quantities are influenced more by increase of temperature and less by change in precipitation. Regional climate model simulations currently in progress driven by global model results will enable us to determine if refinement of dynamical processes, particularly for summertime precipitation, will influence these relationships.