Here, a regional climate model (RCM) RegCM4 and Coupled Model Intercomparison Project phase 5 (CMIP5) global climate models (GCMs) such as Coupled Physical Model (CM3), Coupled Climate Model phase 1 (CM2P1) and Earth System Model (ESM-2M) with their representative concentration pathway (RCP) datasets were utilized in projecting hydro-climatological variables such as precipitation, temperature, and streamflow in Teesta River basin in north Sikkim, eastern Himalaya, India. For downscaling, a ‘predictor selection analysis’ was performed utilizing a statistical downscaling model. The precision and applicability of RCM and GCM datasets were assessed using several statistical evaluation functions. The downscaled temperature and precipitation datasets were used in the Soil and Water Assessment Tool (SWAT) model for projecting the water yield and streamflow. A sequential uncertainty parameter fitting 2 optimization algorithm was used for optimizing the coefficient parameter values. The Mann–Kendall test results showed increasing trend in projected temperature and precipitation for future time. A significant increase in minimum temperature was found for the projected scenarios. The SWAT model-based projected outcomes showed a substantial increase in the streamflow and water yield. The results provide an understanding about the hydro-climatological data uncertainties and future changes associated with hydrological components that could be expected because of climate change.