In this paper, a procedure is introduced to predict the service life of non-return valves using production planning data of injection molding machines. Time and cost pressures are challenging plastic processing companies. Major problems are the frequently used maintenance strategies. Non-return valves are usually replaced following specified maintenance intervals. On the one hand, these components are often carried out too early, not utilizing the full capacity of the components service life. On the other hand, a failure causes time-consuming and costly production downtimes. This paper shows an approach for an automated lifetime prediction of non-return valves by using past production planning data (production data, setting and data records and maintenance logs) from manufacturing execution systems used by plastic processing companies, which are generally stored long-term. Based on these data, load magnitudes, load spectra and the damage accumulation model are created using expert knowledge. Relevant model input parameters (such as processed material, part weight, material throughput) are selected in terms of their availability in the production planning data and their effect on component wear to design load magnitudes. The weighting, and thus the respective influence of these parameters, is done during the calibration with available data. Based on the present production order, a prediction of the remaining service life of the non-return valve carried out. By this, the paper shows the potentials for improving manufacturing execution systems through intelligent use of production planning data for lifetime prediction, which results in the increase in overall equipment effectiveness.