The focus of this study is the quantification of multianalyte mixtures in water by the use of sensor arrays based on polymer layers. Reflectometric interference spectroscopy is used as a optical sensor system for temporal-resolved measurements of the interaction kinetics of analytes in water with the polymer layers. The principles and widespread possibilities of this approach are demonstrated using the quantification of quaternary aqueous mixtures of low alcohols from methanol up to 1-butanol. The sensitive layers consist of two hyperbranched polyesters and one microporous polyimide. Different time-dependent sensor signals are evaluated by artificial neural networks. Because the kinetics of sorption and desorption of the analytes differ significantly, the number of sensors needed for a quantification of analytes in mixtures can be reduced. A feature extraction allows identification of the most important differences of kinetic patterns of the analytes and allows improvement of the multivariate calibration. It is shown that a quantification of quaternary mixtures of methanol, ethanol, 1-propanol, and 1-butanol is possible on the basis of only two polymer sensors.