Advances in DNA related technology hold significant potential for application to biomonitoring but also hold risks and challenges. An important advance has been the adoption of ancient DNA approaches for the detection of environmental DNA (eDNA) - DNA that exists in the environment. All organisms shed DNA and much finds its way into the environment and can be detected by sampling water, soil, air, or faecal material. This provides a non-invasive approach to biodiversity surveillance and holds considerable promise for applications such as the detection of rare species, presence of invasive species in the early phase of the invasion curve, or for whole community inventory. The main risks lie in inadequate delineation of the limits of detection leading to the potential for incorrect diagnoses of species presences or absences and in the inadequate databases and systems to fully interpret the information that emerges from multispecies DNA profiles. In order to overcome some of the implicit challenges, we have developed a framework to estimate the sensitivity of both the field and laboratory components eDNA survey methods, and can demonstrate how these can be used to estimate the overall sensitivity. We have applied this framework to species-specific eDNA surveys to estimate the sensitivity, or probability of detection, for three invasive aquatic species present in Australia; Perca fluviatilis, Cyprinus carpio, and Misgurnus anguillicaudatus. We have also developed a method for the detection of spawning in a threatened species, Macquaria australasica, and how eDNA can also detect the presence of terrestrial vertebrate species at water sources. Examples from each of these applications will be presented and how eDNA can potentially transform species monitoring with the overall caveat of robust, and scientifically defensible interpretation of results.