Recent investments on petrochemical industry of Middle East Asian countries have increased the production of HDPE from cheap ethane, and accordingly, the competitions in this field have been severely intensified. Meanwhile, its market has been extended to various industries, which has led to more frequent grade-change operations to satisfy the diversified customer demands. In this study, an on-line property prediction model has been proposed to predict MI and density of HDPE, which are the most significant properties in its production, based on operating conditions. This could minimize the amount of off-specification products which are inevitable during the grade-change operation. The prediction model is applicable not only to steady-state operations, but also to grade-change operations. Moreover, a dynamic model is proposed on the basis of the develop property prediction model to determine the operating conditions required for optimal grade-change operations. Consequently, the process could be optimized at the whole plant level as well as the local reactor level.