Recently, artificial shelters have been proposed as a novel tool to monitor Hellbenders (Cryptobranchus alleganiensis) and other cryptobranchid salamanders. Factors that influence artificial shelter use by Hellbenders have not been identified, but are important for maximizing the utility of these shelters as monitoring tools. To identify these factors, in 2013–2018 we deployed 438 artificial shelters across 10 stream reaches inhabited by Hellbenders, within three rivers in the upper Tennessee River Basin. We hypothesized that occupancy and nesting would depend on shelter placement, and would be greatest in reaches with relatively high densities of adult/subadult Hellbenders (i.e., >1.5 individuals per 100 m2). We placed shelters in locations representing a range of instream conditions, but avoided microhabitats that were not suitable for Hellbenders. We monitored shelter occupancy by Hellbenders every 2–8 wk, and surveyed shelters for nests every 2–5 d during their breeding season. We quantified densities of adult/subadult Hellbenders and 10 habitat variables across multiple spatial scales. Hellbenders occupied 46% of artificial shelters, and nested in 17% of artificial shelters that were in place for at least one breeding season. Hellbenders were most likely to occupy and nest in shelters placed in portions of those reaches that were ≥50 cm deep with high densities of adult/subadult individuals. Among the variables we considered, population density was the most important factor influencing shelter occupancy by Hellbenders. Shelter nesting was most influenced by water depth, but also by population density and time since shelter installation. Both occupancy and nesting in shelters increased for 2–3 yr following shelter deployment. Our results provide evidence that artificial shelters constitute efficient tools in some streams for monitoring the occurrence and reproduction of Hellbenders.