This paper presents a chemometrics-assisted optimization study to improve the separation of tocopherol (-T) and tocotrienol (-TT) homologues on a C 30 stationary phase in reversed-phase HPLC. The HPLC settings were optimized using a central composite design and the response surface methodology. Flow rate, column temperature, and mobile phase composition were chosen as independent variables. Peak resolution (R s), analysis time (t R), and peak symmetries of the tocopherol isomers were chosen as response variables. Optimum performance in terms of R s was obtained at a flow rate of 0.31 mL min −1 , a temperature of 8.70 °C, and % B content (methyl tert-butyl ether: methanol: water, 80:18:2, v/v/v) in the mobile phase of 38.12%. The analysis of variance and regression analysis gave adjusted R 2 values of 0.9841 for R s , 0.9850 for t R-(α-T), 0.9853 for t R-(β-T), and 0.9204 for the peak symmetry of β-T. This confirms the good agreement of experimental data with predicted values. The close eluting peaks of β-/γ-tocol could be baseline separated at the optimized conditions at a minimized analysis time. Empirical second-order polynomial models were derived that gave statistically high significances (P < 0.0001). Hence, the models can be successfully employed to predict the optimum separation conditions of co-eluting peaks of β-/γ-tocols. The optimized method was successfully applied to determine the individual tocol homologues in various cold pressed edible oils. Total contents ranged from 15 to almost 2600 mg tocol kg −1 oil.