Eight hundred and ninety students at 35 US institutions of higher education participated in a comprehensive study of textbook use and its relation to ratings of textbook quality and helpfulness, a student self-report of learning, student self-report of their deep approach to learning, student perceptions of instructors, and a measure of quiz performance. Intercorrelations between key measures revealed surprising relationships about the influence of these variables on self-reported learning and quiz performance. An analysis of textbook differences revealed some significant differences between the books in terms of quality and helpfulness as well as self-reported learning, but not on quiz performance. The authors identified significant predictors of self-reported learning (deep approach and student perceptions of instructor) and quiz performance (grade point average and textbook helpfulness). These results highlight the complexity of examining student learning and suggest some important variables and problems for future research especially the need for a valid, reliable, measure of learning.How can instructors optimise student learning in the college classroom? A key step towards answering this question involves gaining a thorough knowledge of how students, teachers, and textbooks interact to impact learning. How do students interact with the textbook? What are key student characteristics that predict their learning? The current research was designed to address these questions. Students commonly use the textbook as an aid in studying for exams, but many students do not read the textbook to the extent that they should (Gurung & Martin, 2011). In a comprehensive study conducted in the United States, we assessed how student attitudes (about their textbook, about learning, about study strategies, and about their instructor) are related to student performance (measured with a self-report of deep learning and an actual quiz).