This paper presents an expert mechanism approach to manufacturing performance optimization using simulation as the base tool. The expert mechanism is integrated to the back end of a manufacturing simulator to interpret manufacturing simulation results, assess performance, and then, consistent with set constraints, to effect changes on controllable variables prior to the next run to improve performance. The expert mechanism has a knowledge base that includes proven operations management performance-enhancing methods. In contrast, existing commercial simulation-optimization methods use meta-heuristics in which a near-optimum value is searched from a population of alternative solutions, which can be inefficient in terms of time and cost. The findings of a real case study from a world-class manufacturing company are discussed to demonstrate the expert mechanism and compare it with one of the widely used commercial simulation optimizers.