The present research and development study explores new techniques for developing the aircraft troubleshooting generalists demanded for both turn-of-the-century combat effectiveness and peacetime cost-control. An economical, computer-based apprenticeship program is being developed that uses artificial intelligence to provide the benefits of individual tutoring and expert-guided practice of newly developing skills. The computer realistically depicts both aircraft equipment and flightline maintenance aids during troubleshooting simulations to improve transfer of training from the classroom to the job. Rigorous front-end analyses guided the instructional design. These analyses revealed commonalities in equipment and troubleshooting procedures across different aircraft systems, which form the focus of the skill generalization training. Likewise, distinguishing characteristics of expert troubleshooters were identified and targeted as training outcomes. Finally, a theoretical framework has been developed to guide further efforts for improving generalized maintenance problem solving skills.