The spatial dynamics of zoonotic arthropod-borne viruses is a fashionable though challenging topic. Inter-human local transmission of a given arbovirus during an outbreak and its spread over large distances are considered as key parameters of emergence. Here, we suggest that insular ecosystems provide ideal natural ''laboratory'' conditions to uncouple local transmission from long distance spread, and differentiate these two processes. Due to geographic isolation, often-limited land surface area and relatively homogenous ecosystems, oceanic islands display low species richness and often-high levels of endemism. These aspects provide the means for comprehensive entomological surveys and investigations of original host/pathogen interactions. In addition, islands are interconnected through discrete anthropogenic and non-anthropogenic exchanges: whilst islands maintain a substantial level of human and domestic animal exchange with other neighbouring or distant territories, they also comprise dispersal and migratory pathways of volant organisms (insects, birds and bats). Hence, both anthropogenic and non-anthropogenic exchanges in island systems are easier to identify and investigate than in continuous, continental systems. Finally, island ecosystems tend to be notably simpler, more prone to invasive taxa and, therefore, easier to document the colonization or displacement of vector species. These different aspects are presented and overlaid upon the spread of arboviruses within two distinct insular systems: islands of Polynesia and the south-western Indian Ocean. The former have been repeatedly affected by Dengue fever epidemics, while the latter recently suffered four successive epidemics, probably of east African origin, three of which involved the emerging viruses Chikungunya, Rift Valley and Dengue fever. Here, we review some new insights into arboviral spread and evolution associated with investigations that followed these epidemics, as well as several aspects that