Recent water resource crises underlie the need for more research aimed at understanding how climate change may influence the spatial and temporal trends of hydroclimatic variables. Distributed hydrological models (DHMs) that are capable of projecting water resource variability and heterogeneity are valuable simulation tools. A significant advantage of DHMs lies in their ability to take the spatial distribution of land use and soil type into consideration. This study applied a statistical downscale method to NCEP/NCAR reanalysis data and data provided by two General Circulation Models (GCMs), HadCM3 and ECHAM4, to analyze future daily average temperatures and precipitation at 67 meteorological stations in the upper reaches of Yangtze River Basin (UYRB) under two different greenhouse gas emission scenarios (A2 and B2). Meteorological predictions were then used as inputs in a hydrological model, termed EasyDHM, to project future runoff in the river basin over the 21st century. Results showed that average temperatures are projected to increase in the river basin over the next 90 years. Annual rainfall is likely to decrease, with the most reductions occurring during the summer. However, increases in rainfall are possible during the spring. Total annual runoff of the whole UYRB is projected to decline, but increased runoff in sub-regions including the Jialing River Basin is possible. Comparisons between different elevations suggest that temperature variability will be slightly larger in high elevation regions; however, the fluctuation of precipitation in high elevation regions is predicted to be much smaller than fluctuations at low elevations. Overall, the results from this study suggest that climate change could reduce runoff in the UYRB and exacerbate water supply problems in the region.