The interaction between fluid aptitude and two levels of instructional method (expository and discovery) was investigated. The methods differed in the completeness of the support provided for cognitive processing of procedural knowledge. Four performance outcomes were measured--immediate and delayed recall, and near and far transfer of the procedures learned. It was hypothesized that the discovery instructional method would promote transfer among learners with lower levels of fluid aptitude, but would interfere with the cognitive processing of learners with high fluid aptitude. Data were collected from 80 first-year students (randomly assigned to the two treatment groups) at the Thomond College of Education (Limerick, Ireland). Scores on a paper folding test provided a measure of fluid aptitude. Instruction was provided on procedures for solving a computer-based puzzle that involved changing the color configuration of a three-by-three square matrix. The results of stepwise regression analysis indicate that there was no interaction effect on immediate or delayed recall of procedures, or on far transfer of procedures. However, the discovery instructional method produced significantly more correct solutions to far transfer items than did the expository treatment, regardless of the level of fluid aptitude. Discovery instruction was best for learners with lower fluid aptitude, and expository instruction was best for learners with higher fluid aptitude. (Author/TJH)