The status of many amphibian populations remains unclear due to undetected declines driven by disease and difficulties in obtaining accurate population estimates. Here, we used genetic data (Single Nucleotide Polymorphisms) to investigate Australia’s poorly understood Littlejohn’s treefrog, Litoria littlejohni across its fragmented distribution. We detected five identifiable genetic clusters, with moderate to strong genetic isolation. At a regional scale, population isolation likely of population crashes, resulting in small populations with evidence of founder effects. Moderate genetic isolation was detected among populations on the Woronora Plateau despite short distances between population clusters. Evidence of recent declines was apparent in three populations that had very small effective population size, reduced genetic diversity and high inbreeding values. The rates of inbreeding detected in these populations combined with their small size leave these populations at elevated risk of extinction. The Cordeaux Cluster was identified as the most robust population as it was the largest and most genetically diverse. This study exemplifies the value of employing genetic methods to study rare, cryptic species. Despite low recapture rates using traditional capture-recapture demographic methods, we were able to derive population estimates, describe patterns of gene flow, and demonstrate the need for urgent conservation management.