Military fighter pilots spend most of their working hours performing training missions, simultaneously carried out by multiple pilots with different qualifications. Scheduling these missions is typically done manually, which is a time-consuming task resulting in far from optimal schedules. This study proposes a model that schedules three types of training missions and assigns pilots to these missions for a full-year training program. The proposed methodology consists of a two-stage Mixed Integer Program. Both stages are solved with a Brand and Bound algorithm. Additionally, the second stage implements a Rolling Horizon approach. A case study considering two Royal Netherlands Air Force fighter squadrons is used to test the model. The results show an increase of more than 15% for the average amount of training completed by all pilots, compared to the current scheduling process. The solutions are obtained in a few minutes, demonstrating the suitability to use the model in practice and reduce the time spent on constructing schedules.