Environmental DNA-based biodiversity assessments show promise in generating rapid, repeatable, cost-effective data that could supplement, or even replace, traditional biomonitoring methods reliant upon morphological identifications. Despite growing interest and use of genomic based methods in flowing waters, unanswered questions about how the community data generated by traditional and genomic methods differ remain a hurdle for their widespread use across variable environments. Here, we provide a direct comparison of aquatic invertebrate community data generated by 1) traditional morphological identification of benthic samples, 2) DNA metabarcoding of benthic samples, and 3) metabarcoding of environmental DNA collected in filtered stream water. We collected all three sample types concurrently at ~60 stream reaches in 7 watersheds distributed across the continental US (California, Kansas, Oklahoma, Alabama, Georgia, Tennessee, and Virginia). The three methods produced similar gamma diversity at the family level, while morphological identifications gave greater genus gamma diversity, and DNA-metabarcoding approaches had greater species gamma diversity. Similar trends were observed for alpha diversity, with morphological methods identifying more taxa across all levels except for species, where both genomic approaches yielded more taxa. Compositionally, the three methods were similar at coarse resolutions (order and family), but showed differentiation at finer resolutions (genus and species). DNA metabarcoding methods tended to identify relatively more non-insect taxa, and fewer dipterans than the morphological method, while the relative abundances of Ephemeroptera, Plecoptera, and Trichoptera taxa were more similar. Finally, we explored if similar ecological inferences about local and watershed scale abiotic variables that influence invertebrate community structure could be made using data generated by the three different methods. Our work demonstrates that because genomic and traditional methods capture different levels of the invertebrate community, they can be combined to enable more effective biomonitoring and a wide range of research applications.