Changes in climate pattern and/or land cover may alter hydrologic processes, in turn leading to changes in water quality and quantity that may be detrimental to the environment. Climate change may result in increasing variability of extreme precipitation and other hydrologic extremes, leading to greater and more frequent pulses of contaminants being introduced into receiving waters. Analysis and study of the response of hydrology and water resources to climate change have a very important practical and theoretical significance to understand the influence to water resources' development and utilization, planning and management, ecological balance and environmental protection. Miyun Reservoir is the unique surface water source in Beijing. Non-point source pollution, especially agricultural non-point source pollution has now become the main factor influencing the water quality of the Miyun Reservoir. Therefore, the study of the climate characteristics of the Miyun Reservoir basin and discussing changes of the watershed's non-point source pollution load under the influence of climate change, have a very great practical significance to the formulation of policy on water resources protection. A statistical method was employed to analyze the variation tendency of temperature and precipitation based on the meteorological data (during the years from 1961 to 2000) collected from 4 weather stations within the Miyun Reservoir watershed (northeast of Beijing). And then 25 scenarios of temperature and precipitation were generated by a situational method. The hydrological model HSPF (hydrologic simulation program-fortran) was applied to assess the impact of the climate change on the stream flow and nutrient losses of the catchment under different climate change scenarios. The main results showed that: 1) the influence of change in precipitation on runoff was greater than the temperature; if precipitation increased by 20%, it would lead to an increase of 73.37% for the runoff, and on the contrary, if precipitation decreased by 20%, it would result in an decrease of 56.34% for the runoff; 2) there was a positive relationship between the total nitrogen (TN) and total phosphorus (TP) load and the runoff change. When the precipitation increased by 20%, the TN and TP load could potentially increase by 70.8% and 78.3% respectively; and if the precipitation decreased by 20%, the TN and TP load would decrease by 55.3% and 57.2% respectively; 3) For the Chao River watershed, the runoff in wet season was 3.1 times that in dry season, and similarly, the TN and TP load in wet season were 2.9 and 3.5 times those in dry season; for the Bai River watershed, the runoff produced in wet season was 4.6 times that in dry season, and the TN and TP load generated in wet season were 5.6 and 8.5 times those in dry season. The high-risk areas were mainly concentrated in Chao River watershed and the most of the CSAs were close rivers. There was significant discrepancy of the distribution between TN and TP load in this area. Due to the high population density together with crop planting, livestock and surface mining being main industries, Huairou, Yanqing, Luanping and Miyun County had higher loads than other areas, and the non-point source pollution should be controlled using essential measures.