Two fundamental components of behavioral instruction were investigated: the reported testing feature and absolute performance standards. The component analysis was conducted by offering an undergraduate psychology course simultaneously along two dimensions: grading systems and number of study/test cycles. The 425 college student subjects were randomly assigned either to a norm-referenced or criterion-referenced grading condition and to either a one-test or two-test condition. Three types of educational outcomes were investigated: acquisition of subject matter material, subsequent retention, and student motivation and course evaluation responses. The final test performance of students under the mastery paradigm was found to be superior to that of students operating under the traditional system. Multiple-trial students performed and retained significantly better than did one-test students, regardless of grading standards employed. A mastery learning paradigm fostered higher personal grade aspirations and confidence, as well as a greater sense of system fairness and responsivenes to effort, regardless of self-concept level. The relative influence on performance and motivation of number of trials, grading standards, and self-concept were assessed using path analysis, which confirmed that increased acquisition forms a primary causal link between multiple testing and enhanced motivation, which in turn leads to higher subsequent performance. It is concluded that behavioral instruction is an effective instructional technique and that the repeated test feature is the major influence in mastery-learning superiority, although absolute standards did contribute to positive course evaluations. (SW)