The two most important resources blessed by nature to the mankind are land and water. Undoubtedly, these gifts have to be conserved and maintained with unflinching efforts from every one of us for an effective environmental and ecological balance. The efforts and energy of water resources engineers and conservationists are going in this direction to conserve these precious resources of nature. The present study is an attempt to develop suitable methodology to facilitate decision makers to conserve the resources and also reflects the cause mentioned above has been presented here. The main focus of this study is to identify the critical prone areas for soil erosion and computation of sediment yield in a small basin using Universal Soil Loss Equation and Modified Universal Soil Loss Equation (MUSLE) respectively. The developed model has been applied on Sarada river basin which has a drainage area of 1252.99 km2. This river is located in Andhra Pradesh State (AP), India. The basin has been divided into micro basins for effective estimation and also for precise identification of the areas that are prone to soil erosion. Remote Sensing and Geographic Information Systems tools were used to generate and spatially organize the data that is required for soil erosion modeling. It was found that the micro basins with very severe soil erosion are consisting of hilly areas with high topographic factor and 38.01% of the study area has the rate erosion more than 20 t/ha/year and hence requires an immediate attention from the soil conservation point of view. In this study region, though there is one discharge measuring gauge station available at Anakapalli but there is no sediment yield gauging means available to compute the sediment yield. Therefore, to arrive at the suspended-sediment concentration was a challenge task. In the present study the sediment measurement has been carried out with an instrument (DH-48), sediment sampling equipment as per IS: 4890-1968, has been used. Suspended-sediment samples were collected and sediment yield was arrived at the site by using this instrument. The sediment yield was also computed using MUSLE. Data for this model study has been generated from the samples collected from 28 storm events spread over a time span of 1 year, at the outlet of the basin at Anakapalli for computation of sediment yield. The sediment yield as estimated by MUSLE model has been successfully compared with the sediment yield measured at the outlet of the basin by sediment yield measuring unit and found fairly good correlation between them. Hence the developed methodology will be useful to estimate the sediment yield in the hydrologically similar basins that are not gauged for sediment yield.