Natural and anthropogenic boundaries have been shown to affect population dynamics and population structure for many species with movement patterns at the landscape level. Understanding population boundaries and movement rates in the field for species that are cryptic and occur at low densities is often extremely difficult and logistically prohibitive; however genetic techniques may offer insights that have previously been unattainable. We analysed thirteen microsatellite loci for 739 mountain lions (Puma concolor) using muscle tissue samples from individuals in the Great Basin throughout Nevada and the Sierra Nevada mountain range to test the hypothesis that heterogeneous hunting pressure results in source-sink dynamics at the landscape scale. We used a combination of non-spatial and spatial model-based Bayesian clustering methods to identify genetic populations. We then used a recently developed Bayesian multilocus genotyping method to estimate asymmetrical rates of contemporary movement between those subpopulations and to identify source and sink populations. We identified two populations at the highest level of genetic structuring with a total of five subpopulations in the Great Basin of Nevada and the Sierra Nevada range. Our results suggest that source-sink dynamics occur at landscape scales for wide-ranging species, such as mountain lions, and that source populations may be those that are under relatively less hunting pressure and that occupy refugia.