Spatial data processing especially using GIS techniques and integrating with models has been playing an important role in different watershed hydrological process identification. Multiple source spatial and non-spatial data can be processed and integrated for the analysis of spatio-temporal dimension of a given area. The main objective of the study was to assess land use land cover dynamics for the last 27 years and to identify the impacts on surface runoff, sediment loading and soil erosion to provide basis for future water resource management of Borkena watershed by using AnnAGNPS model. To reach the study objective, Landsat satellite images of 1990 TM, 2010 ETM+ and 2017 ETM+ have been used to produce land use land cover classes and to see the dynamics using Geographical Information System (GIS). The results from land use dynamics shows that from 1990 to 2017 urban and cropland have been increased with 8,621.66 ha (31.36%) and 2,105.2 ha (7.65%), while grazing and forest land have been decreased with 10,248.04ha (37.28%) and 478.81ha (1.74%) respectively. The AnnAGNPS model Calibration was carried out from 1990 to 2005 and validation from 2006 to 2016. The result of sensitivity analysis indicated that the CN was the most sensitive parameter to runoff and peak runoff rate whereas LS and K-factor were for sediment yield following RF. For model calibration, R2 of 0.52, 0.62, 0.75; NSE of 0.87, 0.68, 0.28; RSR of 0.32, 4.58, 0.64; and PBIAS of 17.56%, -23.32% and 12.19% were obtained for surface runoff, peak runoff rate, and sediment load, respectively which was very good agreement with the observed values. Validation results indicated that an R2 of 0.76, 0.87, 0.65; NSE of 0.82, 0.72, 0.92; RSR of 0.46, 0.51, 0.27, and PBIAS of 12.42%, -14.55% and 7.86% for runoff, peak runoff rate, and sediment load, respectively. The BMPs results showed that most of the watershed (97%) experienced insignificant soil loss limit and the conversion of 12.25% of highest eroding cropland cells to forest and all crop land convert to grass land would reduce soil erosion, sediment yield and sediment load by 98 %.