Climate change and recent drought events in many areas of the world are expected to bring challenges to crop production and food security. Ability of field studies to answer these challenges depends on the number of treatments and locations that could be included. Crop growth models, following validation and calibration, offer the opportunity to test response of crops to a large number of treatments using long term weather datasets. Cotton2K is a process based crop growth simulation model that was developed to simulate response of cotton (Gossypium hirsutum L.) to climate conditions and field management in semi-arid regions. The objectives of this study were to investigate the response of cotton to irrigations according to the crop coefficient approach in the Texas Rolling Plains; and (ii) determine the best irrigation deficit rate that increases water use efficiency and optimize yield and profit. Field experiments conducted from 2009 to 2012 on an Abilene clay loam soil (fine, mixed, superactive, thermic pachic Agriustolls) at the Texas AgriLife Research-Chillicothe Research station near Chillicothe, TX (34˚15' N, 99˚ 30' W) were used for model calibration and validation. The studies had different tillage treatments and irrigation regimes based on ET replacement method (45%, 66%, 75%, 100%, and 133% ET).  The simulated lint yields were compared against the observed lint yields to validate the model for the Texas Rolling Plains region. The least squares regression analysis between observed and simulated yields was performed to validate the model. The slope of the regression line 0.93 was not significantly different from 1 (Pr > t = 0.335 @ df= 15) and the intercept of the regression line 100.104 was not significantly different from 0 (Pr > t = 0.203 @ df=15).