<p>Caliber rolling is a crucial steel manufacturing technology with high productivity. There are several process parameters affecting the product quality in caliber rolling such as the billet temperature, the roll temperature, the roll speed, the friction coefficient, the reduction ratio, the roll diameter and so on, but they are conventionally adjusted through the trial-and-error method. As the alternative, computer aided engineering (CAE) is widely used. Design optimization is also one of the useful tools to determine the optimal process parameters in caliber rolling. This paper determines the optimal process parameters in caliber rolling using CAE and design optimization technique. The billet temperature, the roll temperature and the roll speed are the representative controllable parameters, which are handled as the design variables. Considering the characteristics of caliber rolling, the impulse during the process and the distribution of equivalent strain is simultaneously minimized. The numerical simulation in caliber rolling is computationally so expensive that sequential approximate optimization using radial basis function network is adopted to identify the Pareto-frontier between the impulse and the distribution of equivalent strain. It is clarified through the numerical result that the maximum roll force is reduced, and the distribution of equivalent strain is also minimized. It is expected that high product quality material with high productivity will be produced by the proposed approach.</p>