Very often in a manufacturing plant, an important job order must be processed with higher priority than other regular jobs. It is vital for management to understand the impact on production performance of these high priority job orders. In this paper, this dynamic behaviour of assembly line systems is studied using computer simulation and derived metamodels in the form of mathematical expressions. An extensive computer simulation study is first conducted off-line. Next, analytical representations of the system lmetamodels) are fitted to the dynamic behaviour as simulated. These relatively simple equations can then be used on-line to estimate the impact of high priority job order processing when it occurs. The results indicate that the system response to this dynamic event can be described by first-order exponential delay functions captured in metamodels from simulation results. Using only the metamodels, the finish time of high priority jobs can be predicted and the number of delayed regular assembly products can be estimated in real time in the shop floor. This information is very valuable and useful for production scheduling.