We present a general mixed-integer programming model for periodic production scheduling. The formulation, which is based on the state-task network (STN) representation, allows us to model (1) accurate inventory levels, (2) various inventory policies, (3) different demand patterns and profiles, and (4) flexible assignment of tasks to units. We first develop a model for batch processes and extend it to handle continuous processes. The model addresses all the limitations of previous approaches to periodic scheduling and allows us to address various types of periodic scheduling problems. Finally, we demonstrate the performance of the model when applied to different types of processes through several case studies.