Background: The Barotse floodplains of the upper Zambezi River and its tributaries are a highly dynamic environment, with seasonal flooding and transhumance presenting a shifting mosaic of potential larval habitat and human and livestock bloodmeals for malaria vector mosquitoes. However, limited entomological surveillance has been undertaken to characterise the vector community in these floodplains and their environs. Such information is necessary as, despite substantial deployment of insecticide-treated nets (ITNs) and indoor residual spraying (IRS) against Anopheles vectors, malaria transmission persists across Barotseland in Zambia’s Western Province.Methods: Geographically extensive larval surveys were undertaken in two health districts along 102 km of transects, at fine spatial resolution, during a dry season and following the peak of the successive wet season. Larvae were sampled within typical Anopheles flight range of human settlements and identified through genetic sequencing of cytochrome c oxidase I and internal transcribed spacer 2 regions of mitochondrial and nuclear DNA. This facilitated detailed comparison of taxon-specific abundance patterns between ecological zones differentiated by hydrological controls.Results: An unexpected paucity of primary vectors was revealed, with An. gambiae s.l. and An. funestus representing < 2% of 995 sequenced anophelines. Potential secondary vectors predominated in the vector community, primarily An. coustani group species and An. squamosus. Whilst the distribution of An. gambiae s.l. in the study area was highly clustered, secondary vector species were ubiquitous across the landscape in both dry and wet seasons, with some taxon-specific relationships between abundance and ecological zones by season.Conclusions: Larval survey results imply that residual transmission of malaria in Barotseland is being mediated predominantly by secondary vector species, whose known tendencies for crepuscular and outdoor biting renders them largely insensitive to prevalent vector control methods (ITNs and IRS). Alternative interventions may be required against this vector community to further reduce transmission. Larviciding is one potential supplementary intervention, although it presents a challenge in extensive wetlands such as those found in the Upper Zambezi catchment. However, the combination of advances in remote sensing of habitats, genetic identification of local vectors, and spatial modelling present a potential opportunity to optimise targeting of aerial spraying in these environments.