Nonlinear model based control methods are applied to an experimental mini, modular, mobile reverse osmosis desalination system in order to demonstrate the feasibility of the control algorithm for variations in feed water quality and set-point changes. A dynamic nonlinear model incorporating system and membrane properties is derived, fit to the data from the experimental system, and then used to simulate the nonlinear controller's effectiveness and to establish a baseline for comparison to the application to the experimental reverse osmosis desalination system. The nonlinear controller performance is experimentally demonstrated in various setpoint step tests and an integral term is also added to the nonlinear controller formula to eliminate any resulting offset.