We used conventional and spatial analytical tools to characterize patterns of transmission during a community-wide outbreak of dengue fever and dengue hemorrhagic fever in Dhaka, Bangladesh in 2000. A comprehensive household-level mosquito vector survey and interview was conducted to obtain data on mosquito species and breeding as well as illness consistent with dengue. Clusters of dengue illnesses and high-density vector populations were observed in a distinct sector of the city. Dengue clusters are less identifiable in areas further away from major hospitals, suggesting that proximity to hospitals determines whether cases of dengue are diagnosed. Focusing on those areas relatively close to hospitals, we found a spatial association between dengue clusters and vector populations. Households reporting a recent dengue illness were more likely to have Aedes albopictus larvae present in the home when compared with households not reporting cases. Households reporting a recent dengue illness were also more likely to have a neighbor with Ae. albopictus present in the home. In contrast, the presence of Aedes aegypti within the premises as well as the homes of neighbors (within 50 meters) was not associated with dengue illness. Given that the breeding habitats for Ae. albopictus are somewhat distinct from those of Ae. aegypti, the findings of this study have implications for control of dengue transmission in this urban setting where much of the focus has been on indoor mosquito breeding and transmission. Public health officials may find the disease-environment map useful for planning targeted interventions because it displays areas where transmission is most intense.