High quality palm oil biodiesel can be produced by combination of alkali transesterification and separation processes in a TiO2/Al2O3 membrane reactor. Due to the small molecular size, methanol molecules are able to pass through the membrane along with the products. Since methanol is one of the process reactants it is necessary to be recovered. Methanol recovery by means of continuous distillation was employed and the influences of different operational parameters including heating temperature permeate flow rate and reactants ratio were investigated. The results indicate that these parameters have significant effects on the rate of methanol recovery. To simulate the rate of methanol recovery an artificial neural network based approached was employed. The experimental data obtained in this work were used to train and validate the created neural network. (C) 2010 Elsevier B.V. All rights reserved.