[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI AT AUTHOR'S REQUEST.] Terracing is a conservation practice to reduce erosion and intercept runoff from steep lands. Terraces control erosion and runoff by dividing long slopes into shorter slopes; thus, decreasing slope length, which reduces the magnitude and velocity of concentrated flow and allows for sediment to deposit in the cut segment of the terraces. The Soil and Water Assessment Tool (SWAT) is a continuous time, semi-distributed, watershed scale hydrologic model widely used to evaluate runoff and erosion. To account for terrace effects on runoff and erosion, SWAT has relied on reducing the slope length, adjusting the empirical Universal Soil Erosion Equation (USLE) support practice factor (P-factor), and adjusting the hydrologic runoff Curve Number (CN). This tool has limitations, and the runoff and erosion may not be well estimated because of changes in land shape after terrace installation. A modification of the SWAT (called SWAT-Terrace or SWAT-T) explicitly simulates runoff and erosion from terraces by separating terraces into three segments instead of evaluating the entire terrace. SWAT-T aims to improve the simulation of the hydrologic process of runoff and erosion from terraces. The objectives of this work are to 1) evaluate the performance of SWAT-T for simulating the terrace benefits on runoff and erosion from the Goodwater Creek Experimental Watershed (GCEW) at the Hydrologic Response Unit (HRU) and watershed scales, and 2) compare terrace benefits on runoff and erosion estimated with SWAT and with SWAT-T in GCEW. The SWAT model was parameterized for the slope length, USLE P-factor, and the CN. The SWAT-T model was parameterized for slope length, steepness, and USLE P-factor for three terrace segments. Data from 1993-2010 measured at the watershed outlet were used to evaluate the models. To estimate terrace benefits on runoff and erosion, models were compared with and without terraces. Results of SWAT-T showed good performance for the monthly runoff, but poor performance for the monthly erosion. This is probably because of large storms in spring 2002 that prevented planting, causing poorly simulated scheduling of actual field operations. SWAT-T showed [about]2 percent reduction in runoff and [about]13 percent reduction in erosion at the HRU scale and showed 0.1 percent reduction of runoff and [about]3 percent reduction in erosion at the watershed scale. For comparison of terrace benefits on runoff and erosion estimated with SWAT and with SWAT-T, SWAT-T showed more benefit in runoff and erosion at the HRU scale compared to SWAT. Results of SWAT-T showed a 13 percent reduction in runoff and a 95 percent reduction in erosion with terrace installation. Conversely, SWAT showed only a 0.03 percent reduction in runoff and an 89 percent reduction in erosion. Studies using the SWAT-T model indicated that the model may be used for quantifying the terrace benefits on runoff and erosion from terraced-cropped HRUs and watershed scales. Terrace algorithm incorporated in SWAT (SWAT-T) allowed model estimated runoff and erosion trapping in the cut terraced segment leading to better estimation of runoff and erosion.