Active acoustic instruments (echosounders) are well-suited for collecting high-resolution information on fish abundance and distribution in the areas targeted for tidal energy development, which is necessary for understanding the potential risks tidal energy devices pose to fish. However, a large proportion of echosounder data must often be omitted due to high levels of backscatter from air entrained into the water column. To effectively use these instruments at tidal energy sites, we need a better understanding of this data loss and how it may affect estimates of fish abundance and vertical distribution. We examined entrained air contamination in echosounder data from the Fundy Ocean Research Center for Energy (FORCE) tidal energy test site in Minas Passage, Nova Scotia, where current speeds can exceed 5 m·s-1. Entrained air depth was highly variable and increased with current speed, and contamination was lowest during neap tides. The lower 70% of the water column and current speeds <3 m·s-1 were generally well-represented in the dataset. However, under-sampling of the upper water column and faster speeds strongly affected simulated fish abundance estimates, with error highly dependent on the underlying vertical distribution of fish. Complementary sensing technologies, such as acoustic telemetry and optical instruments, could be used concurrently with echosounders to fill gaps in active acoustic datasets and to maximize what can be learned about fish abundance and distribution at tidal energy sites.