Future space missions will be very different from current missions. Mission durations will be significantly longer than current Space Shuttle missions, new systems will be more complex than current systems, and resources will have to be used more efficiently than they are at present. Furthermore, delays in communication between space crews and Earth-based support will necessitate greater crew autonomy than is presently required. To adequately prepare NASA personnel for these challenges, new training approaches, methodologies, and tools are required. This proposal outlines a research program aiming at developing these training capabilities, and builds on significant accomplishments achieved in the past year. Well-designed interfaces, tasks, procedures, and training are critical defense layers in preventing error, and in promoting mission success. They are also critical for the early recognition of errors once made, and for minimizing the consequences of errors. Thorough understanding of human cognition, learning, and skill acquisition are foundational ingredients in the proper design process. As such, research in learning not only contributes to the design of training programs, but also to the design of the systems and the procedures to be trained. Because validating training implementations and particularly those aimed at the long-term retention of skills takes time, this research must commence as soon as possible so as to have finalized products in time to meet the needs of the Constellation Program. What’s more, intermediate products from this research effort benefit current missions and allow iterative improvement cycles with continuous feedback from key stakeholders. With sufficient time for iterative cycles of development, improvements in current training programs could lead to significant improvements in future systems design. This opportunity to contribute to system design is the result of the fact that training programs must often compensate for design deficiencies.