A challenge in managing vector-borne zoonotic diseases in human and wildlife populations is predicting where epidemics or epizootics are likely to occur, and this requires knowing in part the likelihood of infected insect vectors dispersing pathogens from existing infection foci to novel areas. We measured prevalence of an arbovirus, Buggy Creek virus, in dispersing and resident individuals of its exclusive vector, the ectoparasitic swallow bug (Oeciacus vicarius), that occupies cliff swallow (Petrochelidon pyrrhonota) colonies in western Nebraska. Bugs colonizing new colony sites and immigrating into established colonies by clinging to the swallows’ legs and feet had significantly lower virus prevalence than bugs in established colonies and those that were clustering in established colonies before dispersing. The reduced likelihood of infected bugs dispersing to new colony sites indicates that even heavily infected sites may not always export virus to nearby foci at a high rate. Infected arthropods should not be assumed to exhibit the same dispersal or movement behaviour as uninfected individuals, and these differences in dispersal should perhaps be considered in the epidemiology of vector-borne pathogens such as arboviruses.