The analysis of environmental DNA (eDNA) allows efficient surveys of freshwater species and is being increasingly used. However, most studies generally have a limited sampling plan. Because different methods are used by different authors, all the data produced with eDNA cannot be easily pooled together for a global overview. Nevertheless, one of the promising perspectives is the standardization of the methods and protocols, for long-term monitoring, early detection of invasive species, and rare species detection. We here present the biggest dataset using eDNA metabarcoding with standardized methods for freshwater bivalves. Sampling was performed mainly in France, with over 350 localities spread over all major basins. Metabarcoding was performed using two new sets of primers, one for the Unionida and one for the Venerida. The species distributions inferred from eDNA sampling are then compared with those currently known. Results give an insight into the valuable data that eDNA analysis can provide if used at a large scale. Some species were found outside of their known range, especially invasive species but also species of conservation interest. Others were found in a surprisingly low number of localities, although currently considered widespread and of no conservation interest. This can be explained by the databases currently used to establish conservation status, which are often biased by including ancient data, shell-only data, and potentially misidentified species. The present study shows that our eDNA metabarcoding approach will be an essential tool to unveil the true species distributions, to better evaluate their status and improve conservation policymaking.