Dengue occurred sporadically in Bangladesh from 1964 until a large epidemic in 2000 established the virus. We trace dengue from the time it was first identified in Bangladesh and identify factors favourable to future dengue haemorrhagic fever epidemics. The epidemic in 2000 was likely due to introduction of a dengue virus strain from a nearby endemic country, probably Thailand. Cessation of dichlorodiphenyltrichloroethane (DDT) spraying, climatic, socio-demographic, and lifestyle factors also contributed to epidemic transmission. The largest number of cases was notified in 2002 and since then reported outbreaks have generally declined, although with increased notifications in alternate years. The apparent decline might be partially due to public awareness with consequent reduction in mosquito breeding and increased prevalence of immunity. However, passive hospital-based surveillance has changed with mandatory serological confirmation now required for case reporting. Further, a large number of cases remain undetected because only patients with severe dengue require hospitalisation. Thus, the reduction in notification numbers may be an artefact of the surveillance system. Indeed, population-based serological survey indicates that dengue transmission continues to be common. In the future, the absence of active interventions, unplanned urbanisation, environmental deterioration, increasing population mobility, and economic factors will heighten dengue risk. Projected increases in temperature and rainfall may exacerbate this.