A chemometrics approach has been used to optimize the separation of eight coumarin compounds by mixed micellar liquid chromatography. A utility function, a multi-criterion decision-making (MCDM) method, was tested for evaluation of two different measures of chromatographic performance (resolution and analysis time). The effect of six experimental parameters on a chromatographic response function (CRF) was investigated. The factors studied were the concentrations of SDS and Brij-35, alkyl-chain length of the alcohol used as organic modifier, organic modifier concentration, mobile phase pH, and temperature. The experiments were performed according to a face-centered cube response-surface experimental design. For evaluation of the chromatograms a simple linear response function was used which expressed as a summation of two optimization criteria, resolution and analysis time. Then calculated CRF values were fitted to a polynomial model to correlate the CRF values with the variables and their interactions. The regression model obtained was characterized by both its descriptive and predictive ability (R=0.963 and R2cv=0.861) and used, by means of a grid-search algorithm, to optimize the chromatographic conditions. Experiments performed under the optimum conditions predicted by the model produced a chromatogram of high quality. The model was also verified by the good agreement observed between predicted and experimental values of the chromatographic response function under the optimum conditions.