The enhancement of severe disease upon secondary infection makes dengue almost unique among infectious pathogens and presents a serious challenge to vaccine design. Several key observations have been made which shed light onto this phenomenon particularly that antibodies can enhance Fc receptor-dependent uptake of virus into macrophages thereby increasing virus replication. Furthermore there seems to be a relationship between the peak virus load and disease severity. However, a second key feature of dengue is that the life-threatening symptoms do not correlate with the period of high viraemia; instead they occur at a time when the virus load is in steep decline. The coincidence of severe disease manifestations with defervescence and virus control suggests that the symptoms may be a consequence of the immune response to the virus rather than virus induced cytopathology. One of the key elements in the immune response to viruses are T cells which can both secrete a host of inflammatory cytokines and also be directly cytotoxic to infected cells. There are a number of experimental models of T cell-induced immunopathology including in responses to viruses. Particularly interesting in this respect are models of RSV-induced immunopathology, which have direct relevance to vaccine design as a formalin-inactivated vaccine to RSV actually enhanced disease in children when they became naturally infected with RSV, an echo of the disease enhancement seen in dengue. We will present an analysis of CD8+ T cell responses to a number of novel T cell epitopes during dengue infection and also analyse the function and cytokine secretion of these cells. We suggest that an exaggerated and partially misdirected T cell response seen in secondary dengue infection may be part of the complex series of events leading to dengue haemorrhagic fever and shock.