The Council of Independent Colleges (CIC) has long been concerned about the educational success of students, particularly underrepresented groups of students. As part of its Creating Pathways to Educational and Economic Opportunity in Urban Colleges and Universities project (the Pathways Project), CIC organized 19 institutions, nine in urban and ten in non-urban areas, to explore students' performance on the Collegiate Learning Assessment (CLA). In addition to collecting data from a representative sample of students, participating institutions agreed to draw an in-depth sample of first-generation students (defined based on parental education) and low-income students (defined based on Pell-grant eligibility). The in-depth sample increased the sample size of underrepresented groups and thus allowed for more accurate estimates of their CLA performance. The data included in this report are cross-sectional, with institutions collecting information from samples of first-year students in fall 2010 and seniors in spring 2011. The first portion of this report focuses on CLA performance of underrepresented groups of students. Though descriptive results reveal gaps in CLA performance between underrepresented groups and their more advantaged peers, those differences can be accounted for by student characteristics. Thus, after adjusting for student characteristics, particularly academic preparation, no notable gaps in CLA performance appear among different groups of students. More specifically: (1) Descriptive results indicate that first-generation students perform less well on the CLA than non-first-generation students; (2) Similarly, after adjusting for individual-level characteristics, there are no differences in CLA performance in either first or senior year between Pell-eligible and non-Pell-eligible students; and (3) Descriptive gaps in CLA performance by race/ethnicity are substantially larger than those by first-generation status and Pell-grant eligibility. The second portion of the report focuses explicitly on CLA performance of both students and institutions in urban and non-urban settings. These analyses reveal several key findings: (1) Student-level analyses reveal no difference in CLA performance in urban and non-urban settings, after adjusting for students' individual-level characteristics. One exception is the performance task of the CLA in the senior year, when students in urban contexts seem to perform less well than students in non-urban settings; (2) Institutional value-added analyses confirm results obtained from individual-level models. On average, urban institutions have slightly lower value-added scores than non-urban institutions with respect to the performance task measure, but not with respect to other components of the CLA; and (3) Institutional value-added analyses also reveal large variation among institutions within both urban and non-urban contexts. There are institutions in each context that have positive and negative value-added scores. This variation "within" each setting by far overshadows any overall differences between urban and non-urban settings. These findings have several notable implications: (1) The most important predictor of CLA performance is academic preparation; (2) Urban and non-urban institutions on average perform equally well on the CLA, and most of the variation is within specific institutional settings (i.e., urban and non-urban) as opposed to across them; and (3) CLA performance varies more within institutions than across them. All institutions have students who perform at different levels, producing much more variation within institutions than across them. (Contains 22 figures, 16 tables, and 3 footnotes.)