The traditional way of marking student’s assignments is considered exacting and tedious, particularly when managing a large number of students’ assignments. Likewise, it makes the task marks prone to being lost and difficult to oversee. To address this issue, this paper proposes an automatic marking system based on optical character recognition (OCR) to mark the students’ answers automatically with better accuracy that will save teachers’ time and effort, facilitate the process of assigning marks, and tackle the problem of marks being lost. Moreover, this paper proposes a novel measurement to effectively calculate the final mark of students’ submitted answers. The proposed system consists of two main phases to achieve the main goal, which are: (i) image processing with OCR, and (ii) matching the results with optimal answers. The proposed system is evaluated using a real dataset, where the experimental results showed that the proposed system achieved high efficiency and more accuracy by using such a simple yet effective technique in automatic marking.