Boulder reefs are common habitats in the northern hemisphere, characterized by ecologically distinct hard bottom underwater structures, acknowledged to be hot spots of biodiversity providing refuge for benthic organisms and fish communities. Monitoring of the biota associated with these reefs is currently done by divers equipped with underwater video and radio communication. While this approach is robust, it is of interest to explore the usefulness of modern molecular techniques that are less expensive, scalable, and may provide more detailed information on plants, animals and microorganisms associated with the reefs. We investigated the use of eDNA metabarcoding for supplementing traditional diver-based monitoring of biodiversity of marine boulder reefs within the photic zone. We applied a sampling design, which made it possible to evaluate the usefulness of eDNA monitoring as a supplement for traditional monitoring. Specifically, we 1) assessed the local imprint of boulder reefs on biodiversity across the North Sea to Baltic Sea transition zone and 2) investigated the importance of environmental gradients for patterns in community structure. We compared the composition and abundance of species associated with nine reefs, representing an environmental gradient of salinity (16 to 33 psu), water temperature (16 to 21°C) and water depth (6 to 29 m). At each reef location, water was sampled near the bottom, over the reef and ca. 2.6 km upstream and downstream (placement) and sequenced with metabarcoding using COI, 18S and 12S primers. eDNA identified 400 species, diver based observations identified 184 with an overlap of 70 species (12%) and 81 genera (18%). While eDNA identified many infaunal species, it failed to detect several macroalgal species which dominated in the diver based observations. Multivariate analysis of both eDNA and diver derived community structure distinguished between reef communities, with a significant match (r = 0.37, p = 0.02). Furthermore, the eDNA approach made it possible to identify significant differences in species composition between upstream, reef and downstream sites, suggesting that eDNA leaves a local footprint in benthic habitats. Patterns in both eDNA and diver based species composition and richness were significantly related to geographical distance, salinity, water temperature and water depth. Despite of low detection of macroalgae, the eDNA sampling provided a substantial supplement to traditional diver based monitoring of biodiversity around benthic hotspots in the Danish marine waters and therefore we recommend to add eDNA methods to conventional monitoring programs in the future.