In June of 2003, the U.S. Supreme Court upheld the University of Michigan Law School's practice of considering race in admissions by a margin of 5-4 (Grutter v. Bollinger), but struck down the formulaic approach for admitting freshman by a margin of 6-3 (Gratz v. Bollinger). Even though the Court narrowed the use of race by rejecting mechanical scoring systems that assign bonus points to underrepresented students, it left the door open for colleges and universities to continue to consider race in admissions to enroll a critical mass of racially/ethnically diverse students. Among the members of the Court itself, there was major disagreement over the value and effects of diversity in an educational setting. For example, Justice Sandra Day O'Connor, who authored the majority opinion in Grutter, wrote that body diversity is a compelling state interest that can justify using race in university admissions, whereas her counterpart Justice Antonin Scalia wrote in the dissenting opinion that he was not convinced that educational benefits flowed from diversity and listed this shortcoming among a long list of other issues that can potentially bring about future lawsuits. Given the conflicting opinions, it is clear that the controversy regarding claims about the educational benefits of diversity is far from settled, and there continues to be a pressing need to understand empirically how students actually benefit, if at all, from being in more racially/ethnically diverse environments. This study applies a multilevel approach to examine the student- and institution-level effects of one key form of diversity--namely, frequency of cross-racial interaction. Two key research questions guided this study: (a) How do college students who report high levels versus low levels of cross-racial interaction compare with regard to the educational outcomes of openness to diversity, cognitive development, and self-confidence? (b) How do students who attend institutions with high peer versus low peer average levels of cross-racial interaction compare on measures of openness to diversity, cognitive development, and self-confidence? Background Because of the recent national attention on the constitutionality of race-conscious admissions practices, a growing body of empirical research about diversity has emerged in the last 10 years, focusing mainly on racial/ethnic diversity with particular interest in enrolling a larger proportion of underrepresented students (African American, Latino/a, & Native American). Several publications have recently reviewed this body of research (see, for example, Chang, Witt, Jones, & Hakuta, 2003; Hurtado, Dey, Gurin, & Gurin, 2003; Hurtado, Milem, Clayton-Pedersen, & Allen, 1998, 1999; Milem & Hakuta, 2000; Smith, Gerbick, Figueroa, Watkins, Levitan, Moore, et al., 1997). Basically, these reviews showed that diversity-related benefits are far ranging, spanning from benefits to individual students and the institutions in which they enroll, to private enterprise, the economy, and the broader society. There was remarkable consistency among these reviews concerning both the empirical studies they considered and the conclusions they drew. One important conclusion that emerged from these reviews is that the vitality, stimulation, and educational potential of an institution are directly related to the composition of its student body, faculty, and staff. A number of studies have shown that campus communities that are more racially diverse tend to create more richly varied educational experiences that help students learn and prepare them better for participation in a democratic society (Antonio, 2001b; Astin, 1993a; Bowen & Bok, 1998; Chang, 1999; Chang et al., 2003; Gurin, Dey, Hurtado, & Gurin, 2002; Hurtado, 2001; Milem, 1994; Orfield & Kurlaender, 2001; Pascarella, Edison, Nora, Hagedorn, & Terenzini, 1996; Sax & Astin, 1997). One reason for this appears to be that race still shapes opportunities and experiences in U. …