Soil loss by rainfall/runoff is danger and continuous ecological problem in Gilgel Beles watershed. Over the years various soil erosion researches have been conducted in upper Abbay river basin. These researches reason out for the crises of soil erosion has been driven on failure of land use land cover. But to know the significant effect of sub -catchment slope steepness, rain fall magnitude and land management system on soil loss, this study was conducted to estimate potential annual soil loss rate and to evaluate area prioritization for conservation planning at Gilgel Beles watershed. Revised universal soil loss equation (RUSLE) based spatial soil erosion modeling and Multi criteria decision evaluation technique (MCDE) method integrated with ArcGIS were considered in Gilgel Beles watershed to estimate potential soil loss and to give area priority for better conservation planning. For gully erosion susceptiblity maping using pairwise comparison and analytical hierarchy process model was implemented. The annual soil loss rates of the watershed ranges from 0 to 511.2ton/ha/yr with average annual soil loss rate of 28.68 ton ha-1 yr-1. The total annual soil loss was 2.21Millon ton yr-1. The spatial pattern of extremely severe, severe, very high and high, soil erosion contributed were 1.42 %, 3.55 %, 5.79% and 12.81% of the watershed area respectively. However, 76.43% of the watershed area soil erosion rate ranged from moderate to low. Factors contributed for severed soil erosion includes slope steepness, land cover change to farm, heavy rain fall intensity and poor land management system. The MCDE result revealeds that, among the various gully erosion factors, upslope steepness was rank first and contributed significant influence on the gully erosion occurrence at lowland of the watershed. Based on the result of MCDE map, the saturated lowland areas of Dangila and Mandura districts have been provided very high gully erosion. But results, obtained using RUSLE presented higher erosion rates mostly was in the upper sections of the watershed. The study showed that the initiation and expansion of gully erosion in the watershed is facilitated by (rain fall intensity, topography and soil texture). So the study indicated that the combined models were effective for soil erosion estimation and to identify gully erosion in the watershed and it was proved that team work have needed to apply effective soil erosion control techniques for erosion prone areas. Unless investigation tells soil erosion can affect the study watershed.