Monitoring biodiversity changes over large spatiotemporal scales is critical for effective ecosystem conservation and management. This study investigates the potential of environmental DNA (eDNA) metabarcoding to enhance national-scale biomonitoring of freshwater diversity by leveraging discarded filters associated with routine water quality sampling from the U.S. Geological Survey's (USGS) National Water Quality Network (NWQN). We tested 375 samples from 103 NWQN sites for eDNA of native and non-native fish and found that 52% of the filters yielded fish eDNA for a total of 70 fish species detections. Of the filters that had fish eDNA present, an average of 3.7 species were detected. Benchmarking these results to USGS's Aquatic Gap Analysis Project (AGAP)—which includes both field-verified observations along with predictive models derived from fish capture and landscape predictor datasets—we found that eDNA from these filters detected only a fraction of the observed and expected fish diversity for these sites. Our results indicate that these discarded filters may not be sufficient for eDNA sampling of fish communities and posit that alternative filter types more appropriate for eDNA sampling may yield more valuable biomonitoring data. Nevertheless, we tested the efficacy of two novel approaches to facilitate large-scale biomonitoring. Though these filters did not yield adequate fish eDNA, the AGAP database provides a useful method for ground truthing fish species presence. The potential of integrating eDNA sampling into existing monitoring frameworks, which, when paired with more optimal eDNA methods, could be a cost-effective strategy to enhance biodiversity monitoring at large scales.