A membrane transport model suitable for the multiple solutes system in reverse osmosis is developed for unsteady-state simulation and prediction of membrane filtration dynamics in terms of permeate flux and concentration of each solute. This model is based on coupling the concentration polarization model using unsteady-state differential material balance and extended Spiegler-Kedem model. This model is characterized by the parameters solute diffusivity in the concentration polarization layer (D-bi), reflection coefficient (cri), osmotic constant (a(i)), hydraulic permeability constant (LP), mass transfer coefficient (k(i)) and solute permeability coefficient (P-ii). These parameters are estimated by using the Levenberg-Marquardt method coupled with the Gauss-Newton algorithm using the experimental data. The experimental data were obtained from the treatment of pretreated palm oil mill effluent (POME) as a feed in the pilot plant scale reverse osmosis system. The pretreated POME composed of a ternary system with the solutes of carbohydrate constituents, protein and ammoniacal nitrogen. The simulation results show a good agreement with the experimental data. The proposed model is suitable for predicting the performance of multiple solutes in a reverse osmosis process. The concentration of each solute present is correlated with the COD of the permeate stream. (c) 2006 Elsevier B.V. All rights reserved.