Mesopelagic fishes are an important component of the world’s oceans in terms of their abundance, biomass, and ecosystem function. These fishes are important contributors to the biological carbon pump via their feeding and behaviors, whereby they facilitate the transfer of carbon from shallow waters to the deep sea. Several species undertake diel vertical migration, feeding in shallower waters at night and moving to deeper waters during the day. This process actively expedites the downward flux of carbon. However, carbon budgets and climate models require accurate information regarding the depth distributions and migration patterns of these fishes, and environmental DNA (eDNA) analyses can provide this information. Here, we utilize eDNA approaches, generating taxonomically-informative COI and 12S reference barcodes for 80 species of mesopelagic fishes, which can be used for species-level identification of eDNA sequences. Using these, along with a publicly available barcodes database, we compare results from eDNA analysis with traditional net sampling, and explore the ability of eDNA techniques to detect diel vertical migration in fishes from samples collected in Northwest Atlantic Slope Water. We found that eDNA and net samples often resulted in different species identifications, demonstrating that eDNA can detect species that would otherwise be missed with traditional methods. In our eDNA samples, we also detected more species (12) in our shallowest depth category (0 - 100 m) from night samples than from day samples (3). This is consistent with increased diversity in shallow waters at night due to diel vertical migration. Based on the variability observed in sample duplicates, we suggest that future mesopelagic eDNA studies incorporate larger sample volumes and scaled-up sampling efforts. We also note the potential applications of eDNA analysis in addressing ecological questions related to predator-prey relationships identification of foraging hotspots, and carbon flow through the ocean’s midwaters.