A Soil and Water Assessment Tool (SWAT) model was built to assess non-point source (NPS) pollution loads in the Ashi River basin in China. Future climate data from the Hadley Centre's coupled ocean/atmosphere climate model (HadCM3) in Representative Concentration Pathways (RCP) 4.5 scenario were downscaled. These downscaled data were then used as an input to simulate changes in future NPS pollutant loads. The model estimated NPS-total nitrogen (TN) and NPS-total phosphorus (TP) loads at 2507.78 tons per annum (t/a) and 429.26 t/a, respectively, by 2014. In the RCP4.5 scenario, NPS loads were predicted to show an increasing trend in the future. Assuming no changes in land-use in the future, NPS-TN and NPS-TP pollution rates showed increases of 8.8% and 7.9%, respectively, in 2099. The risk of NPS pollution in Ashi River Basin was greatest for Rice Land (RICE), followed in descending order by Agriculture Land (AGRL), Urban Land (URBN), Pasture (PAST), and then Forest Land (FRST). The study concludes that measures such as fertilizer reductions and establishing vegetated filter strips should be introduced to reduce the NPS pollution. When model scenario was set fertilizer usage reduced by 10%, the NPS pollution output of TN and TP was reduced by 5.9% and 2.8%, respectively. The area of the filter strip was proportional to the pollutant removal rate; the reduction effect was better for NPS-TP than NPS-TN. The results of this study can help water resource managers implement better management strategies in the Ashi River basin.