The purposes of this study were to examine whether ethnic differences in college students' self-construals exist and whether these self-construals are associated with preferences for different types of classroom organization; and also to test the validity of self-construal theory to explain psychological differences. Self-reports were collected from 197 American college students (White = 87, Hispanic = 56, Black = 54) in three different types of higher education institutions in the Southwest region of the U.S. A series of MANOVA or MANCOVA models indicated that degree of identification with one's ethnic group (covariate), rather than the independent variable of ethnic group, was crucial in predicting one's independent and interdependent self-construal scores. No ethnic differences existed for any of the preferences of classroom organization. Additionally, a multiple regression mediation analysis indicated self-construals appeared not to act as mediators of students' preferences for learning organizations, as opposing to the self-construal theory. Further research is needed to test the validity of self-construal theory or the self-construal scale. (Contains 3 tables.)