Student engagement has become problematic following the rise of mass and universal forms of higher education. Significant attention has been devoted to identifying factors that are associated with higher levels of engagement, but it remains the case that the underlying reasons for student engagement and, indeed, the notion itself of 'student engagement' remain weakly theorised. In this article, we seek to develop the theoretical basis for student engagement in a way that highlights the student's own contribution. We explore how learning involves students taking responsibility for action in the face of uncertainty, whether in pursuit of personal or communal concerns. Drawing on perspectives primarily from realist social theory, we suggest that student engagement may be shaped by extended, restricted and fractured modes of reflexivity and co-reflexivity. In this way student engagement in higher education is theorised as a form of distributed agency, with the impact of a learning environment on this agency mediated by reflexivity. Reflexivity itself is further influenced by the tasks and social relations encountered by students in a given learning environment. The role that social relations play in students' responses to learning specifically offers a means to strengthen the moral basis for education. Our account provides an explanation as to why specific educational practices, such as those termed 'high impact', might lead to higher levels of student engagement within the wider context of a knowledge society. We thus offer insights towards new forms of educational practice and relations that have the potential to engage students more fully. education institutions evidently stand to gain a good deal from any capacity to foster an engaged student body.Research has also begun to identify specific educational practices that are particularly effective in engaging students. Studies using data from the National Survey of Student Engagement (NSSE) in the US, for instance, have been able to establish a set of practices that have a high impact on student engagement (Kuh & Schneider, 2008). While such an approach may assist in selecting educational practices, though, it leaves the underlying concept 'student engagement' relatively weakly theorised. Kuh and Schneider (2008), for instance, identified ways in which the level of academic challenge, the presence of active and collaborative learning, student-faculty interaction and a supportive campus environment contribute to student engagement, but without seeking to theorise how these elements work together to effect gains of various kinds. Such studies tend to emphasise the way in which student engagement is linked to given institutional or student characteristics, but Carini et al. (2006, p. 23) emphasises that only a relatively modest proportion of the variation in learning outcomes can be accounted for through measures of student engagement based on survey instruments. Coates (2006, p. 57) has argued that this framework from Kuh and Schneider (2008) is the most advanced exi...