Microemulsion EKC (MEEKC) was used for the determination of ketorolac and its three impurities. The microemulsion system was optimized, for the first time in the literature, using a multivariate strategy involving a mixture design. A 13-run experimental plan covering an experimental domain defined by the components aqueous phase (10 mM borate buffer pH 9.2), oil phase (n-heptane) and surfactant/cosurfactant (SDS/n-butanol) was carried out. Good results were obtained with all microemulsions tested considering as responses analysis time and resolution, and according to the desirability function the best microemulsion system was constituted by 90.0% 10 mM borate buffer, 2.0% n-heptane, 8.0% of SDS/n-butanol in 1:2 ratio. Finally, with the aim of reducing analysis time, a response surface study was carried out in the experimental domain defined by the process variables temperature and voltage and the best values were 17 degrees C and -17 kV, respectively. Applying the optimised conditions, a complete resolution among the analytes was obtained in about 3 min using the short-end injection method. The method was validated for both drug substances and drug product and was applied to the quality control of ketorolac in coated tablets. A comparison of MEEKC, MEKC and CEC for assaying ketorolac and its related substances has been made.