To determine the impact of college experiences, it is critical to have measures of student learning and development that are valid, reliable, and affordable. Data on student outcomes are useful not only to institutions that want to assess their students' growth but also to external constituencies, such as federal and state governments, that demand evidence about the effects of the college experience. Having accurate and affordable measures of student development enables researchers to identify aspects of the college experience that are most educationally effective. This study focuses on differences in the correspondence between self-reported and longitudinal gains across race and ethnicity, gender, and social class. In addition, because some evidence suggests that students with greater cognitive ability are more accurate in assessing their own test performance, potential differences across academic achievement are also explored. This study includes both cognitive and noncognitive outcomes. Some of the longitudinal gains were assessed with objective measures (exams), whereas others used subjective measures (students' perceptions of their attributes). The breadth of the types of outcomes and the forms of longitudinal assessments enhance the generalizability of the findings. (Contains 7 tables.)