This study investigates the use of metacognitive prompts and training structure for a computer-based training intervention. It is framed in resource allocation theory (Kanfer & Ackerman, 1989) and includes a fully-crossed 2 (low vs. high structure) x 2 (no prompts vs. metacognitive prompts) experimental design to examine how different training methods affect training outcomes – performance and self-efficacy – via metacognitive activity. Individual differences such as cognitive ability, motivation, and goal orientation were also anticipated to affect trainee performance on a test of immediate and delayed performance. Results from this research indicate that structure affected scores on a delayed performance test but not immediate performance or self-efficacy. Furthermore, metacognitive prompts did not produce any predicted effects on performance. Future research should carefully consider the viability of metacognitive prompts for affecting training outcomes.