ABSTRACT
The literature on class size and student performance is reviewed. Student performance and class size in a large modular course was studied over a period of 10 years, involving over 5,000 modules and 250,000 student grades. The hypothesis that students would perform less well in larger classes was confirmed. A small but highly significant negative relationship between module enrolment and grades was found and students in larger classes stood significantly lower chances of getting good grades. Subject areas varied widely in the strength of the relationship between class size and performance. As class size increased markedly during this period and performance was worse in larger classes it was hypothesised that overall student performance would have declined during the period studies. This hypothesis was not confirmed: overall student performance improved. Possible explanations of this paradox are explored.