Land cover change strongly affects biodiversity in stream ecosystems, with several studies demonstrating the negative impacts of agricultural and urban expansion on local community richness. However, little is known of the effects of land cover on the variation among sets of local communities in stream networks, as well as the drivers of community variation in these systems. Using the metacommunity framework, this study takes a multi-scale approach to understand how macroinvertebrate communities are assembled across three catchment land cover types; native forest, agricultural and urban. Specifically, the aims of this study are to assess; (1) how stream network land cover influences alpha and beta diversity of macroinvertebrate communities and, (2) the relative role of local environmental conditions and spatial dispersal variables in structuring these communities. Benthic macroinvertebrate samples and local in-stream and riparian environmental variables were collected at 20 sampling sites in each of the six study stream networks in Auckland. Spatial distance proxies of macroinvertebrate dispersal in stream networks were calculated using geospatial techniques. Community alpha and beta diversity, environmental and distance variables were analysed using multivariate statistical techniques. Comparisons showed reference forest and impacted (agricultural and urban) networks supported distinct communities, with lower alpha diversity in the impacted stream networks. Unexpectedly, beta diversity in the impacted networks was greater than, or equal to the reference stream networks, with community dissimilarity almost entirely driven by species turnover. Overall, irrespective of land cover, macroinvertebrate communities were largely structured by local environmental conditions. Benthic substrate and the presence and composition of riparian vegetation were the most significant local environmental variables influencing community composition. Spatial dispersal limitation variables had a small, but significant, effect on inter-site community dissimilarity and overall community structure in each catchment. Network distance between local communities explained the greatest variation in community dissimilarity of the three distance types. This study identified potential drivers of macroinvertebrate community variation in Auckland streams, specifically highlighting the relative role of local environmental and spatial dispersal processes. The results of this study have relevance for biomonitoring and state of environment reporting of Auckland’s freshwater systems, as well as future stream rehabilitation projects.