The throughput of all manufacturing systems is con-strained by one or more bottlenecks. We recently developed a novel method to find the bottlenecks in a manufacturing system and to return a quantitative measure of the bottleneck constraint. This shifting bottleneck detection method has now been expanded to monitor the bottlenecks for non-steady state and highly variable discrete event systems, allowing the user to monitor these running bottlenecks over time as the underlying system changes. This paper presents the method for monitoring these shifting bottlenecks over time. The method is an extension of the highly reliable shifting bottleneck detection method. This method uses a novel holistic approach to the analysis of discrete event system, analyzing not only the entities by itself but also the interactions between the entities. The paper includes an example to demonstrate the method.