Artificial shelters show great promise as novel, non-invasive tools for studying hellbenders, but their use thus far has faced several challenges. During initial trials in multiple river networks, artificial shelters routinely became blocked by sediment and dislodged during high stream discharge events, and were rarely used by hellbenders. We sought to determine whether these complications could be overcome via alternative shelter design, placement, and maintenance. Between 2013 and 2018, we deployed 438 artificial shelters of two different designs across ten stream reaches and three rivers in the upper Tennessee River Basin. We assessed evidence for several hypotheses, postulating broadly that the availability, stability, and use of artificial shelters by hellbenders would depend on how shelters were constructed, deployed, and/or maintained. We found that maintaining shelters at least once every 40 days limited sediment blockage, and building ~ 40 kg shelters with 3-4 cm thick walls and recessed lids improved their stability during high discharge events. Additionally, we found that hellbenders most frequently occupied and nested in artificial shelters when they were deployed in deeper (~50+ cm) portions of reaches with high adult hellbender densities. Our results suggest that artificial shelters can serve as effective tools for studying hellbenders when designed, deployed, and maintained with these advancements, but also highlight some limitations of their use. / Master of Science / Hellbenders are large, fully-aquatic salamanders that live primarily in cool, rocky, swift-flowing streams in portions of Appalachia and the lower Midwest. They are imperiled across most of their native range due to human-caused habitat degradation, but their declines, conservation needs, and population status have historically been difficult to study due to the fact that they spend the majority of their lives beneath large, often inaccessible boulders. While these boulders are sometimes possible to lift, doing so can disturb critical hellbender habitat. Therefore, alternate, less invasive hellbender sampling methods are necessary in order to improve knowledge about their conservation status and needs. Artificial shelters, which are large, hollow, concrete structures that mimic natural boulder crevices and feature removable lids, show promise as a novel, innovative tool for non-invasively studying hellbenders. However, initial trials of these shelters have yielded mixed results, with shelters often becoming swept away and destroyed during floods, becoming blocked by sand and sediment and thus inaccessible to hellbenders, or simply not being used by hellbenders when accessible. We sought to determine whether these complications could be overcome by optimizing the way that shelters were constructed, deployed, and maintained in streams inhabited by hellbenders. Between 2013 and 2018, we deployed 438 artificial shelters of two different designs across ten stream reaches and three rivers in the upper Tennessee River Basin. Using multiple analyses, we tested one broad overall hypothesis: that the efficacy of using artificial shelters to study hellbenders would depend on how they were constructed, how frequently they were maintained, and where they were placed in the stream. We found that maintaining shelters at least once every 40 days limited sediment blockage, and building ~ 40 kg shelters with 3-4 cm thick walls and recessed lids improved their stability during flood events. Additionally, we found that hellbenders most frequently occupied and nested in artificial shelters when they were deployed in deeper (~50+ cm) portions of reaches with high adult hellbender densities. Our results suggest that artificial shelters are effective tools for studying hellbenders when designed optimally, maintained frequently enough, and placed in appropriate locations. However, exceptions to these findings may exist in certain heavily degraded stream reaches.