General Certificates of Secondary Education, known as GCSEs, are organised into modules which can be assessed at the end of the course (linear route) or at different points throughout the course (modular route). This research investigates students' performance in English and mathematics in each assessment route and the effects of certification session and re-sits on examination outcomes using data from one English awarding body. In mathematics, modular students performed better than linear students once prior attainment was accounted for. In contrast, modular assessment in English did not always lead to better performance. In the modular routes of English, students certificating early in the course were at a disadvantage compared to those certificating at the end. However, modular mathematics students obtained significantly better results in early sessions than at the end of the course. The benefits of re-sitting seemed clear, as most students performed better on their second attempt at a module than on their first. However, this did not often lead to an improvement on the module grade. (Contains 7 tables, 2 figures, and 5 notes.)