Previous literature calls for further investigation in terms of precedents and consequences of learning approaches (deep learning and surface learning). Motivation as precedent and time spent and academic performance as consequences are addressed in this paper. The study is administered in a first-year undergraduate course. Results show that the accounting students have a slightly higher score for deep learning compared to surface learning. Moreover, high intrinsic motivation and extrinsic motivation have a significant positive influence on deep learning. Next, deep learning leads to higher academic performance; surface learning on the other hand leads to lower academic performance. The effect of deep learning on performance still holds, when we control for time spent, gender and ability. Consequently we can conclude that a deep learning approach is much more than "simply" spending a lot of time on studying.