Understanding the intricate dynamics of biodiversity within and across riverine ecosystems, influenced by geological history and environmental factors, is crucial for effective conservation and management strategies. Italy, particularly the Ligurian region, harbors diverse freshwater fish communities and populations shaped by unique geological and hydrological conditions. Here, we investigated the suitability of eDNA-metabarcoding to identify inter- and intraspecific diversity patterns of riverine fish populations along the main drainage divide (MDD) between the Adriatic and Ligurian basins in Northern Italy. We collected 96 aquatic eDNA samples across 48 riverine sites, amplified them using a cytb primer pair and denoised the sequences to retrieve amplicon sequence variants (ASV). We calculated communities’ phylogenetic distance with betaMPD based on genetic distances derived from the ASVs, combined them with conductance-based landscape metrics and applied generalized dissimilarity models (GDM) to assess spatial genetic structure. Our results reveal genetic differentiation among populations of several fish species, with some displaying clustering patterns across the drainage divide and isolation by distance patterns. Overall, taxon richness was significantly higher in the Ligurian sites (25) than in the Adriatic side of the MDD (22), as was ASV richness (205 vs. 196). Our findings highlight the effectiveness of eDNA-metabarcoding in uncovering various facets of diversity, shedding light on hidden genetic diversity within ASVs, and revealing significant spatial genetic structuring in freshwater fish populations across multiple species.