Modern flight training devices often require performance validation against actual aircraft flight test data to help produce realistic performance and handling qualities. Prior to performing the trainer validation task, a significant analysis and manipulation effort is required to develop the flight test data into a complete and consistent set of test criteria. By performing this comprehensive data analysis early in the program development, flight modeling validation problems can be identified, minimizing the risk of cost and schedule overruns. This paper addresses the data analysis and development process performed on a modern helicopter flight trainer (AH-1W) using off-line software analysis tools, simulation modeling feedback, and extensive customer interaction. Off-line software tools are used to rapidly and efficiently perform such tasks as identifying and resolving discrepancies in the test data base, performing polynomial curve fits and data extrapolations, normalizing similar data sets, and graphically comparing data acquired from different maneuvers and from different aircraft. The paper addresses how the trainer flight simulation model can be used to adjust or establish trends in the data or to resolve conflicts between similar data sets from different sources. The necessity of extensive customer involvement in this iterative test criteria definition process is stressed. The discussion concludes with specific recommendations on the data acquisition and analysis process, based on lessons learned, including the application of trainer specification tolerances.