Dams and weirs are a significant source of environmental degradation in rivers. One hypothesis explaining how they reduce species diversity is that dams and weirs are barriers (i.e. breaks in connectivity) to dispersal along rivers. Such connectivity can be pivotal to population persistence and also to restoration of degraded reaches. However, it is unknown whether and how dams or weirs block or impede dispersing species, hence diminishing the likelihood of successful downstream restoration. The aim of this study was to test whether and how a weir on a tributary of a regulated river impeded dispersal of stream insects. First, in order to understand the effects of a weir and its associated pool, we determined how drifting insects were affected by natural, slow-moving pools. Insect drift entering and exiting 13 natural pools was measured twice, autumn and early summer. Second, we measured drift entering and exiting a weir when the weir was over-topping. We then tested whether drift rates through natural pools differed from drift through a weir. We found the weir reduced drift to a greater extent than natural pools for three of five study species (Offadens spp., Austrophlebioides spp., Austrosimulium spp.), and the caddisfly, Cheumatopsyche sp. AV4, was significantly impeded only during early summer. The reduction was potentially caused by the interaction of drifting insects with the unique hydraulic character of the weir pool, where the water depth at the downstream end of the pool decreased abruptly due to the weir wall. This substantial alteration of flow profiles may have limited movement through the weir pool to insects drifting near the water surface. The obstruction of drift by the weir suggests that any attempts at downstream river restoration may become predominantly reliant on dispersal of flying adults from populations elsewhere within the river system.