Terrestrial salamanders (Plethodontidae) are increasingly cited as important surrogates for monitoring forest ecosystem health and biodiversity. To accurately assess the ecological role of salamanders, total abundance is more relevant than the population subset available for sampling. Therefore, estimating detectability is important for generating unbiased population metrics, as abundance estimates average an order of magnitude greater than surface counts. The past two decades have seen a tremendous increase in analytical methods that account for imperfect detection. Terrestrial salamanders have unique physiological requirements and surface activity patterns that contribute to their inherently low detection levels, and variable terminology makes it difficult to compare detectability parameter values across species, regions, methods, and studies. It is important to use consistent terminology and explicitly state what specific components of detectability are being reported in a study. Availability is governed by the presence of individuals in the sampling area and the capacity of the search method to detect the targeted animals. Conditional capture probability is affected by survey methodology, observer experience level, habitat complexity, and species crypsis. Imperfect detection affects all population-related metrics, including abundance, occupancy, and species richness. Akin to capture–mark–recapture methods, population analysis via hierarchical modeling is a recently developed approach that can account for imperfect detection. Hierarchical models enable researchers to partition complex ecological systems into simpler component models and focus on the mechanisms underlying the observed data. With improved population analyses of terrestrial salamanders, monitoring and conservation efforts can become more precise and effective.