Simulation-based expert systems have successfully been used in technical troubleshooting training systems for many years. There are two parts to a simulation-based expert system: the simulation and the advisor. The simulation contains knowledge about a technical system. The advisor is a logical inference engine that processes information in the simulation to determine where tests should be made to isolate a fault in the system. When simulation-based expert systems are used for training, they provide advice to students who need help isolating a simulated malfunction. Advances have been made in the sophistication of such simulation-based expert systems over time. However, with advances come tradeoffs for development. This poster will review the features of early, intermediate and recent simulation-based expert systems developed for technical training. This review provides the foundation for discussing some of the key issues involved in the development of such systems. The issues presented include those for both knowledge elicitation and knowledge representation. Sample SBES systems will be demonstrated.