Production-inventory systems model the interaction of manufacturing processes with internal and external customers. The role of inventory in these systems is to buffer mismatches between production and demand caused by process uncertainty. Often, production and demand variability is described using simplified probabilistic models that ignore underlying characteristics such as skewness or autocorrelation. These models lead to suboptimal inventory policies that result in higher costs. This work presents a novel analysis of the impact of uncertainty in the performance of production-inventory systems. It quantifies the effect of different probabilistic descriptions of production capacity and demand in systems subject to lost sales or backorders. The analysis is based on the results of discrete-event simulations. The flexibility offered by simulation allows studying diverse conditions that arise in production-inventory systems. The results clearly illustrate the importance of appropriately quantifying variability and performance for inventory management in process networks.