This paper evaluates the capability of the reflectance-based crop coefficient coupled with a water balance model to estimate crop evapotranspiration (ET) and soil water depletion on rain-fed grapes. A grape orchard located in southeast Spain was studied during the 2009 growing season. Model results were evaluated by comparison with ET measured by a Bowen Ratio and Eddy Covariance systems and soil water content measured with a Capacitance Probe system. Comparison between measured and modelled ET showed a good agreement (RMSE = 0.53 mm d-1) and the greatest difference between measured and simulated water content was 7% in relative error. By testing the model on this type of fruit tree and taking into account the results, the methodology using visible and near infrared remote sensing to obtain a reflectance-based crop coefficient gains robustness and confidence for future operational applications.