Abstract This article investigates the effectiveness of two distinct formative assessment methods for promoting deep learning and hence improving the performance amongst engineering students. The first method, applied for undergraduate students, employs a lecturer-led approach whereas the second method uses a student-led approach and e-learning for postgraduate teaching. Both studies demonstrate that the formative assessment and feedback has a positive effect on the performance of engineering students, especially those lying on the middle and lower grade tail. The mean exam marks increased by 15 to 20% as a result of introducing formative assessment to the case study modules. The main catalysts for performance improvement were found to be the feedback provided by the lecturer to the students, and by the students to their peer partners. Comparison of the two practices leads to the conclusion that whilst both methods are equally effective, peer assessment requires less time commitment from the lecturer.