To ascertain changes in cognitive correlates of learning as students advance through hierarchical instruction, 24 individual difference measures were obtained from 166 Navy trainees who had completed a computer-managed course in electricity and electronics. Principal component analysis and varimax rotation were computed for cognitive characteristics, producing factor scores that were used in multiple regression analyses to predict achievement in 11 modules of instruction. During acquisition of course content, the cognitive components sampled shifted noticeably in importance throughout the curriculum. The results have implications for research on aptitude treatment interaction (ATI), transition from novice to expert, crystallized and fluid intelligence, task demands of instruction, and computer managed mastery learning. (Author/LMM)