Connectivity and dispersal are key components determining the persistence of fragmented populations. However, the assessment of dispersal rates and landscape factors affecting functional connectivity remains problematic for species living in small populations, even more so in stress-sensitive, rare or elusive species. In this study, we estimated current dispersal patterns in a regional population of the endangered capercaillie (Tetrao urogallus) using a novel, parsimony-based application of pedigree analysis. In spring 2003 and 2008, we systematically collected non-invasive samples of this forest bird species in five local populations in the north-eastern Pre-Alps of Switzerland. Based on 11 nuclear microsatellites, we identified 86 unique genotypes over the two sampling periods and determined levels of genetic structure. We investigated the relative impact of topography and land use on functional population connectivity by comparing inter-individual genetic distance with GIS-derived least-cost path models representing different dispersal scenarios. Results show that local populations are connected by dispersers, suggesting occasional contemporary gene flow. Mountain ridges were found to hamper capercaillie dispersal, while, at odds with our hypothesis, neither a broad valley nor land use type notably affected gene flow. These results suggest that our study region represents a patchy metapopulation. Our study illustrates how pedigree analysis and landscape genetic methods can significantly contribute to the understanding of dispersal and connectivity in rare and elusive species.