The main thrust of this paper was to examine students performance in different class sizes in a hypermedia learning environment. Participants in the study were one hundred and fifty six (156) students of junior secondary schools, selected using purposive sampling method. Hypermedia treatments were administered in the different class sizes, with pre and post tests, which lasted 6 weeks. The study found that students exposed to hypermedia instruction in both small and large classes performed significantly better in their post tests. It was also found that the variables sex, class size and modes of hypermedia instruction have significant interaction on the performance of learners. The study concluded that class size variable should be considered in planning hypermedia instruction.