Soil erosion or sedimentation is a major problem of reservoir operation in Ethiopia. Deforestation, overgrazing and poor land management practice are some which accelerate the rate of erosion. The topography of Ethiopia in general is full of ups and downs and local farmers commonly cultivate on the hilly sides causing easy topsoil wash away. Hence, this study has tried to determine the sediment yield at the dire dam reservoir, identify the high sediment source sub basin and check the applicability of SWAT model on the dire dam watershed. To go through this objective, SWAT model was applied with methodology of collecting hydro metrological data, sediment data, topographic, land use and soil map data by overlaying mechanism, the SWAT model run. Simulation was carried out using metrological and spatial data by dividing the watershed in to 9 sub basins with 103 hydrological response unit (HRUs) at outlet of dire watershed. Based on the availability of data the model calibration period (1988-1999) and validation period (2000-2006) were performed for monthly flow and sediment data using Sequential Uncertainty Fitting (SUFI-2) with in SWAT calibration uncertainty program (SWAT-CUP). The model was found applicable in this watershed with the performance evaluation statistics (Nash-Sutcliffe model efficiency (ENS), coefficient of determination (R2)) in the acceptable range, (R2in the range of 0.74 to 0.78, ENS in the range of 0.73 to 0.77). From the model simulation output, sub basin 1, 2, and 6 were found the top three severely eroded sub basin with average annual sediment yield of 19.01 t/ha,34.26t/ha and 27.31t/ha respectively. While sub basin 7 and 9 were found the least sediment source sub basin with annual average sediment yield of 8.06t/ha and 7.22t/ha respectively. Generally, the annual averaged sediment inflow in to dire dam reservoir was 1850 t/km2/yr. Three watershed management scenarios were simulated to compare their effectiveness of sediment yield reduction from the existing baseline condition. The result shows that the mean annual sediment yield at the outlet can be reduced by 34.14% by applying filter strips, 57.12% by applying grassed waterway, and 70.06% by applying terracing. To generalize, applying terracing is relatively effective to reduce the mean annual sediment yield for the proposed dam.