Hydrologic and water quality investigations are fundamental to any watershed management programme. The need for accurate measurement of parameters involved in hydrological and environmental processes has grown rapidly due to the acceleration in the watershed management programmes for resource conservation, development and assessment of nonpoint source pollution pertaining to soil and water resources. Surface hydrologic modelling of watershed mainly includes processes like runoff and transport of sediment as well as pollutants from the watershed. Though various methods have been suggested at international level and many models are being utilized, but there is not much work carried out in Indian conditions and that to especially in the state of Chhattisgarh. The recent version of Soil and Water Assessment Tool (SWAT) interfaced with ArcGIS and called as ArcSWAT, being used widely and successfully in different countries including India is adopted for simulating the hydrological behaviour for watershed management in upper Mahanadi basin. ArcSWAT uses a two-level disaggregation scheme; A preliminary sub-basin identification is carried out based on topographic criteria, followed by further discretization using land use, soil type and slope considerations. Areas with the same soil type, slope and land use forms a Hydrologic Response Unit (HRU). Mahanadi is one of the major inter-state east flowing perennial rivers in peninsular India. Andhiarkhore gauging station of Central Water Commission (CWC), located at the outlet of Hamp watershed of Seonath Sub-basin of upper Mahanadi basin was considered for this study. The study area lies between 810 01’ E to 810 36’ E and 210 45’ N to 220 30’N with an altitude ranges from 267-977 m above mean sea level (MSL). The total catchment area of Hamp river is 2210 sqkm. The current study was under taken with the use of a distributed parameter model interfaced with GIS i.e. ArcSWAT to estimate the surface runoff, sediment yield and nutrient losses from a critical agricultural sub-watershed of Hamp watershed for development of effective management plan. The study area was subdivided into 14 sub-watershed considering topographical parameters derived from DEM and drainage network. The land cover, soil layers, and DEM were used to generate 207 HRUs to analysis of daily and monthly seasonal precipitation, runoff, sediment yield and nutrient losses, for 2004 - 2008 (calibration period) and 2010 - 2013 (validation period) Simulated daily and monthly runoff and sediment yield for the monsoon period for both calibration and validation period compared well with their observed counterparts. Capability of the model for generating rainfall was also evaluated for the period of 2004 - 2008. The model simulated daily rainfall was having close agreement with that of observed rainfall (R2 = 0.638, ENS = 0.677 and PBIAS = 5.102). Also the model predicted daily and monthly runoff and sediment yield using generated daily rainfall compared well with observed runoff and sediment yield during simulation period of 2004 - 2008. The effect of storage structure in a sub-watershed on runoff and sediment yield at the watershed outlet was also evaluated. Input variables used for model calibration were soil conservation service (SCS) curve number, base flow alpha factor, ground water delay time, threshold depth in shallow aquifer, plant uptake compensation factor, soil evaporation compensation factor, effective hydraulic conductivity in main channel alluvium, Manning’s “n” value for the main channel. Sensitivity analysis was performed to find the effect of these parameters on runoff and sediment yield. Sensitivity analysis was performed using a combined method of Latin Hypercube (LH) sampling and One-Factor-At-a-Time (OAT). Due to non availability of requisite observed data during 2009, the model was validated for the duration 2010 to 2013 using daily rainfall, runoff, sediment yield and seasonal nutrient quality. The time series of the observed and simulated daily and monthly sediment yield of the Hamp watershed for the validation period of 2010 to 2013 were compared graphically which showed that the time to peak of predicted sediment yield matched consistently well with the measured sediment yield throughout the season. The effect of storage structure on runoff and sediment yield was evaluated. The critical sub-watersheds were identified and prioritized based on the sediment yield rates and nutrient loss. Out of these sub-watersheds numbers (WS3, WS6, WS9, WS12, WS13, WS14) fell under moderate soil loss group of soil erosion classes (5 to 10 t/ha/yr) while sub-watersheds (WS4, WS8, WS10 and WS11) fell under high soil loss group of soil erosion classes (10 to 20 t/ha/yr), whereas other sub-watersheds fell under slight erosion classes. Looking at the annual soil loss, runoff and nutrient losses, the most critical sub-watershed was identified as WS4 with runoff (245.97 mm), Sediment Yield (18.18 t/ha), N03-N (1.62 Kg/ha). The prioritization was also supported by morphometric analysis of the Hamp watershed, which shows that WS4 sub-watershed falls under high priority of soil loss. Adequately calibrated and validated ArcSWAT model was used to develop the best management plan for the critical HRUs of the critical sub-watershed (WS4). The sub watershed (WS4) comprises of 15 HRUs (36 - 50) with soybean, maize, rice and sugarcane as its crop. Results showed that rice crop can not be replaced by other chosen crops (maize, sugarcane and soybean) since these crops resulted in higher sediment yield as compared to rice. Soybean and maize crops HRU generates significant amount of soil loss and were considered as the critical HRUs. In order to evolve an appropriate management strategy suited to the farmers of the watershed, a set of 72 combinations, comprising of four crops, six tillage treatments and three fertilizer levels were studied. Futuristic best management practices were developed for 2016 - 2020 (5 years) based on the weather generated data from the calibrated and validated ArcSWAT model for the critical HRUs. In case of soybean crop HRU, the Disk plough (T6) was noted to have considerable impact on sediment yield with 10.8 % increase as compared to the conventional tillage (T5) irrespective of the fertilizer level, whereas in maize crop HRU, M.B. Plough (T5) yielded 3.06% more sediment. The reduction in sediment yield was obtained from conservation tillage (T2) and zero tillage (T1) with 15.31% and 5.68% for soybean and maize crop HRUs, respectively. These results showed that the tillage having higher mixing efficiency produced higher sediment yield in the critical sub-watershed. Treatment of tillage with field cultivator reduces the soil loss by 7.9% with increase in the yield making it sustainable tillage practices to be adopted for both soybean and maize HRUs. Engineering measures were also proposed for arresting soil loss by reducing the runoff volume. Based on Weighted Overlay Analysis (WOA) of various thematic maps, the identification of suitable sites for construction of water harvesting and conservation structures was suggested. Total of 42, 59, 20 and 05 numbers of percolation tanks, small, medium and large check dams were proposed along with suitable sites for arresting erosive runoff and reduction of soil loss.