While many studies have considered the ability of eDNA to assess animal communities in lacustrine settings, fewer have considered riverine systems, particularly those spanning the environmental gradients present in large river basins. Such dynamic systems are challenging for eDNA biomonitoring due to differing eDNA transport distances in rivers and the effects of river chemistry. To address this challenge, we focused on the Thames River system, UK, which has exceptional historical fish records providing a baseline to test the accuracy of eDNA metabarcoding in recovering fish community structure across both fresh and tidal zones. Two primer sets targeting 12S and CO1 regions were used to capture fish communities across the Thames catchment, from the upper freshwaters to the mid estuary. eDNA was collected at 35 sites, 14 of which were simultaneously paired with traditional fish surveys for direct comparison. We demonstrated that eDNA metabarcoding consistently detected more freshwater species than traditional methods, despite extensive sampling effort using the latter. In contrast, metabarcoding did not perform as well as traditional approaches in estuarine waters, although results included the novel detection of the protected sea lamprey. We further demonstrated that minor variations in the recovery of all approaches would not impact on the assessment of simple ecological models of community structure and, thus, some variability between approaches should not be viewed as a serious hindrance to uptake. Rather, our findings support a growing consensus that eDNA can reliably detect fish communities across dynamic freshwater habitats.