Since the inception of off-line quality control, it has drawn a particular attention from research community and it has been implemented in a wide variety of industries mainly due to its extensive applicability to numerous real situations. Emphasizing design issues rather than control issues related to manufacturing processes, off-line quality control has been recognized as a cost-effective approach to quality improvement. It mainly consists of three design stages: system design, parameter design, and tolerance design which are implemented in a sequential manner. Utilizing experimental designs and optimization techniques, off-line quality control is aimed at achieving product performance insensitive to external noises by reducing process variability. In spite of its conceptual soundness and practical significance, however, off-line quality control has also been criticized to a great extent due to its heuristic nature of investigation. In addition, it has also been pointed out that the process optimization procedures are inefficient. To enhance the current practice of off-line quality control, this study proposes an integrated optimization model by utilizing a well-established statistical tool, so called response surface methodology (RSM), and a tolerance-cost relationship.