Creating jet engine simulations that replicate the behavior of actual engine parameters at finite flight conditions is only one step toward meeting the requirements for pilot training. Reproducing realistic performance trends throughout the flight envelope and generating proper responses to malfunctions and pilot-initiated events, including secondary and cascading effects, is critical to achieving positive pilot training. Traditionally, jet engine simulation for pilot training purposes is based on table-lookup of steady-state engine parameters, such as rotor rotational speed, fuel flow, exhaust gas temperature, engine pressure ratio and net thrust. This approach does not inherently meet all the aforementioned requirements and exhibits the following shortcomings: The dynamic engine performance has to be approximated as a lagged transition between steady-state points. It is unreliable to predict the behavior of the engine parameters when excursions outside the bounds of the tables take place. Malfunction effects have to be programmed individually for each engine parameter and for different flight and operational conditions. Additionally, the interdependencies between the different engine parameters can be violated during the model tuning process. Accordingly, a new approach to model jet engines is needed. The objective of this paper is to present a physics-based jet engine simulation approach which addresses the shortcomings of table-lookup solutions, is data-driven and generic, while also distinguishing itself from other physics-based simulations (Claus, Townsend, 2010) by being computationally efficient. This approach can be used to simulate any turbojet or turbofan engine by accounting for the physical processes and the geometric and mechanical characteristics that govern the performance and behavior of the engine. These include the fan, compressors and turbines maps, the rotors inertia, and the thermodynamics of the flow entering the engine from its free-stream state ahead of the engine intake, through the intake duct, the fan, the compressors, the combustion chamber, the turbines and the nozzles. The paper discusses the methodology used in applying the physics-based approach to simulate a two-spool turbofan engine, the technical challenges involved and demonstrates how this new approach advantageously compares with a table-lookup model in matching actual flight test data and in providing realistic performance trends. The paper also assesses the physics-based approach's ability to meet the requirements of the different levels of flight simulators and flight training devices, as defined in FAR 14 CFR Part 60.