Eighty-three (83) high school chemistry students were administered tests of cognitive reasoning level, cognitive restructuring ability, disembedding ability, working memory capacity, and prior knowledge before a learning segment on balancing chemical equations by inspection. After a four-day instructional segment utilizing direct teaching methodology, participants were given a posttest on balancing equations. Initial regression analysis indicated that a multicollinearity problem existed. Factor analysis and correlational data indicated that the reasoning, restructuring, and disembedding variables could be collapsed and redefined as a single restructuring variable. A hierarchial regression analysis was then performed, and the following conclusions were derived: (1) when prior knowledge alone is considered, students' understanding of chemical formulas significantly (p < 0.05) influences overall equation balancing performance; (2) when prior knowledge, restructuring, and working memory are considered, only restructuring ability significantly (p < 0.05) influences overall performance; (3) working memory capacity does not significantly (p < 0.05) influence overall performance but does on certain posttest items; (4) prior knowledge and restructuring ability also significantly (p < 0.05) influence performance on certain posttest items. Discussion includes the rationale for identifying the collapsed variable as restructuring and the absence of working memory capacity as a significant influence on overall performance.