Transformational training solutions like the mixed reality Infantry Immersion Trainer (IIT) are being developed to demonstrate how to create an expeditionary mindset in small units. By framing the training experience to support a range of skills used to make novel and non-routine decisions under simulated conditions that produce stress, time pressure, uncertainty and high risk, the U.S. Marine Corps is transforming its approach to small unit training. The purpose of this paper is to report on applied research for creating training designs that address the shortfalls in the current systems approach to training. These limitations include the inability to rapidly generate training scenarios, lack of standardization in cognitive skills training, mistaking technology for training, and inadequate assessment and feedback tools. Each of these limitations affects training effectiveness. The Future Immersive Training Environment (FITE) was used to demonstrate how to overcome these limitations as part of a Joint Capabilities Technology Demonstration (JCTD), which was sponsored by the U. S. Joint Forces Command (USJFCOM). The key breakthrough was not the technology, though new training technologies were demonstrated. The revolutionary step was to create training designs to practice, develop, and assess small unit decision-making instead of trying to improve procedures and battle drills. The FITE JCTD was a first of kind decision skills trainer where small units could encounter non-routine tactical situations where they were required to make decisions and take actions that would shape their mental models. Case-based training was developed based on the results of cognitive task analysis, which produced a descriptive model of decision skills and their application against the insurgent planning cycle. A family of validated situational judgment test items was administered pre- and post-training to assess whether training had occurred and to capture what parts of the individual mental models had been encoded through effective training.