In this study, a mathematical model is developed for predicting the effect of cosolvents on the biodiesel production, by taking into account the fact that the process consists of two liquid phases. The proposed model consists of a detailed chemical kinetic reaction mechanism and the phase equilibrium of the two phases. The experimental data for estimating the reaction rate constants consists of 1580 dynamic concentration data related to cosolvent-free systems. Then, the calculated reaction rate constants are used to model the biodiesel production process in systems containing cosolvents to investigate the cosolvent effects on the process. The average absolute relative deviation of 1.0% for all the experimental data points is a testament to the accuracy of the predictive model. Hence, the model was used to investigate the effect of methanol to oil ratio, the ratio of cosolvent to methanol, and the amount of catalyst on the biodiesel production rate and yield, in the presence of cosolvents tetrahydrofuran and diethyl ether. According to the results, the use of cosolvent leads to increase the reaction rates. In the other words, the process time is reduced; however, the final yield of biodiesel does not change. Comparison between two cosolvents shows that diethyl ether is more appropriate than tetrahydrofuran, because, in the use of diethyl ether, less cosolvent to methanol ratio is needed. Also, the effect of catalyst concentration is similar to the effect cosolvent. (C) 2016 Elsevier Ltd. All rights reserved.