In many countries, irrigation water is one of the major contributors to water scarcity. In the present study, a novel optimized irrigation system which minimizes water consumption in irrigation is presented. The system is based on a predictive control algorithm, which foresees the water need of the crop, and regulates the time and amount of irrigation to maintain the soil moisture around an optimal level, while taking into account system constraints. The predictive feature of the algorithm requires a model of the soil moisture, which is obtained from the actual meteorological data of the Kingdom of Bahrain. The optimization problem is formulated as an Economic Model Predictive Control (EMPC) problem and implemented using MATLAB. The simulation experiments show that the novel system yields a reduction of water consumption around 8% and 16% compared with the PID and On-off controllers, respectively, while maintaining an optimal soil moisture level.