The relationship between test performance and the size of the class in which instruction occurred was studied at Brigham Young University. The classes prepared students for examinations to meet general education graduation requirements. All test scores from 318 sections representing 27 courses from the fall semester, 1980 were analyzed. Section sizes ranged from 13 to 1,008 students. Sixteen content areas were represented. The findings suggest that at the college level, class size may be a much less important influence on student achievement than previously thought. It appears that increasing class size from current levels of 20 to 40 students up to several hundreds of students may not radically affect college student achievement. It is concluded that decisions concerning class size in colleges and universities should be based on issues other than concern about student achievement (e.g., classroom facilities, course budgets, students' demand for classes). Appendices contain summary statistics for the regression analysis results. (SW)