Following previous research which showed the significance of learners' self-efficacy and dispositions towards progression in school and university transition, we developed and validated similar measures for use in modelling undergraduate students' Learning Gain (LG). We validated three dimensions of confidence using data from a sample of (mainly first year) undergraduate students in various departments of one major UK University which we call: 'confidence in learning through…' (a) 'traditional university transmission pedagogy' (e.g. in lectures), (b) 'social means' (e.g. working in teams), and (c) 'problem solving, reflection and critical thinking'. We explored psychometric properties of these measures and then focus on their association with other perceptions of students' experience in HE and other measures of attainment. We then used these academic self-efficacy variables to inform (regression) models of LG. Finally, we discuss the prospects for measurements and modelling of LG involving dispositions and affect as well as attainment.