Although recent results from the National Survey of Student Engagement (NSSE) suggest there is considerable variation in levels of undergraduate student engagement within Carnegie institutional classifications, research also suggests different classifications of institutions exhibit different types (Pike and Kuh, 2005) and levels (Pike, Kuh and Gonyea, 2003) of student engagement. However, an important question that requires additional examination is the potential relationship between particular engagement variables and various outcomes of interest to students and institutions. Given the time and resource limitations that faculty, staff, and administrators face in their efforts to improve academic quality, analyses that move beyond norm-based or statistical comparisons and test empirical relationships within theory-based models of quality in undergraduate education may provide a more effective means to focus improvement efforts. This study examines student-reported learning and development factor scores, the likelihood of persistence, and students' ratings of their overall educational experience based on NSSE responses from over 1,500 first-year and senior students in 2004 at a large, research-extensive university. Using Chickering and Gamson's (1987) "Seven Principles for Effective Practice in Undergraduate Education" as a conceptual framework to guide variable selection, the results of four models suggest that a high level of academic challenge, institutional emphasis on academic work and studying, and prompt feedback on academic performance may provide the most productive avenues for enhancing the impact and quality of undergraduate education at large, research-extensive universities. Possible explanations and potential implications of these and other findings for students, faculty members, institutional leaders, policymakers, and accrediting bodies are discussed. (Contains 7 tables and 1 figure.)