The quantification of spatial and temporal variability of cotton yield provides critical information for optimizing resources, especially water. The Southern High Plains (SHP) of Texas is a major cotton (Gossypium hirsutum L.) production region with diminishing water supply. The objective of this study was to predict cotton yield variability using soil properties and topographic attributes. The DSSAT CROPGRO-Cotton model was used to simulate cotton growth, development and yield using weather, soil and crop management data. Cotton yield simulations were calibrated using two years (2000-2001) of cotton yield monitor data. The soil data were obtained from the SSURGO database and crop management. Topographic attributes were derived from digital elevation data acquire using a real-time kinematic (RTK) GPS receiver. The prediction of cotton yield variability in relation to soil and topography helps to identify the factors influencing crop yield to make appropriate management decision to increase yield and profits.