Computer-based assessment (CBA) is an important area of e-learning research. Most studies of CBA technology add new constructs to existing user acceptance models and rarely consider the moderating effects. However, the self-regulation (SR) levels (i.e., high or low) in an e-learning environment substantially affect individual learning behaviour and performance. This study investigates the moderating effects of SR levels on the relationships between factors in the CBA acceptance model, which was based on the unified theory of acceptance and use of technology (UTAUT) model. The main findings are that, firstly, both the perceived performance expectancy and social influence significantly affect CBA behavioural intention in all students regardless of their SR level. Secondly, the effect of effort expectancy on CBA behavioural intention is significantly larger in low-SR students compared to high-SR students. Finally, behavioural intention significantly predicts CBA use behaviour in high-SR students but not in low-SR students.