The degree of fidelity required in simulators to effectively transfer newly acquired skills between the classroom and the work world remains illusive and ill-defined during the front end analysis of system design. Frequently, fidelity specifications are inconsistent between the ultimate users of the system, the acquisition agency, and the contractor charged with the design and production of the final training system. Such a situation is not in the best interest of the student and is likely to produce a device insensitive to the directions provided by sound instructional and engineering analyses. This paper presents a technique for allowing individual training tasks to define specific degrees of simulator fidelity and then objectively tracking the task/fidelity relationship throughout the design, development, and testing phases.