A main reverse osmosis (RO) limitation is the operational costs (including energy filtration costs), which are significantly affected by energy costs and the expense of membrane maintenance. Both costs are affected by influent characteristics, including feedwater temperatures (T), electrical conductivity (EC), and recovery set points. A fuzzy logic advanced control system optimizes the RO operational costs considering feedwater EC and T as input values, and the RO recovery setpoint as a control action. Verification results show that, once calibrated, the simulation of the control system can save more than one million euros per year in large RO facilities (>30.000m³·day^-1). Which represents a reduction of 0.11€·m^-3 of influent treated water.