The first phase of the study was a 2 x 2 factorial design, with locus of control and instructional method (lecture and demonstration) as independent variables and honor point average (HPA) as the dependent variable. The second phase used correlational techniques to test the extent to which reading performance and traditional predictors of achievement would predict college HPA. Subjects were marginal admission students at Wayne State University who were economically disadvantaged and 91 percent of whom were black. Forty-three subjects were used in the experimental phase, and 184 were used in the correlational phase. The experimental subjects participated in a reading and study skills course which used either the lecture method or the demonstration method of instruction, and the subjects were divided into groups of internal or external locus of control. The Rotter Internal-External Locus of Control Scale and the Triggs' Diagnostic Reading Test were administered at the end of the reading course to all marginal admission students. After 2 full years of study, HPA's were collected from university records. No significant effects of locus of control, instructional method, or instructor on HPA were found. However, a larger proportion of external locus of control students chose to leave the university than the internal locus of control students. Tables and references are included. (AW)