Condition Monitoring and Predictive Maintenance are one of the areas of the fourth industrial revolution. By accurately estimating the remaining useful lifetime of machine components, the optimal usage of machinery can be guaranteed. Within the context of injection molding machines, screw wear is a factor that not only influences machinery usage, but also strongly affects material quality and the molding process in general. In this talk we describe a method to monitor the state of the screw, by measuring the actual flow rate within the molding process. By comparing this flow rate to a reference flow rate of a new screw, the current state of the screw can be determined. Determining the exact reference flow is the crucial part during this process. Existing models for calculating the theoretical throughput of extruders could be used for its calculation, but they depend on data, which might not always be available. In this talk, the influence of various environmental and process variables is explained and a theory for screw wear is presented. The validity of this theory is shown for real world data.