This study examined characteristics of the student bodies of highly selective and less-selective colleges and universities specifically those related to race and socioeconomic status (SES) and the possible effects of replacing race and ethnicity with SES as a method for ensuring student diversity. The study used data from the Cooperative Institutional Research Program (CIRP), specifically data from freshmen (n=4,408) entering college in 1986. Independent variables included socioeconomic factors (income, parents' educational levels and occupations), race, and background factors (Scholastic Aptitude Test (SAT) scores, high school grades, gender, and high school rank). Institutional selectivity was the dependent variable. Analysis confirmed the hypothesis that academically and socioeconomically advantaged students are more likely to attend highly selective institutions than other students, and six of the nine independent variables were found to be significant predictors of attendance at a selective institution. Although SAT scores and high school grades were the strongest predictors of institutional selectivity, family income, parents' education, and gender also contributed significantly. Also, twice as many white and Asian students attend highly selective institutions as compared to other racial groups. Results suggest that replacing race with socioeconomic status in affirmative action policies for college admissions would assist in diversifying student bodies. (Contains 30 references.) (DB)