Dynamic rate-based approach for modeling and simulating batch distillation is presented. To predict separation efficiencies, mass transfer was considered explicitly using the Maxwell-Stefan. Process dynamics were modeled by considering all relevant dynamic changes in the system, including the column periphery. The model, implemented at the large-scale {ABACUSS} modeling environment, addresses implications of the dynamic and multicomponent nature of the process for its mathematical representation, the use of mass-transfer coefficients, and the need for further experimental correlations. To validate the model, a series of separations of the highly nonideal quaternary system of methyl acetate, methanol, acetic acid, and water in a pilot-plant batch-distillation column with structured packings were conducted. The results show that a rate-based approach can predict the column operation within the experimental error over the entire operation time without fitting of any kind. For the first time the use of a rate-based approach on the ground of the Maxwell-Stefan equations for a multicomponent batch-distillation process was validated experimentally. Comparison of predictions of the rate-based approach with simulations based on an equally tailored equilibrium-stage model shows that the largest differences between the two approaches occur in short periods of very significant changes in the column profile. The additional effort of rate-based modeling can be justified for the design of operating policies of multicomponent batch-distillation processes.