Over the past decade the community-based natural resources management (CBNRM) model has become the trend in African conservation, yet its relative benefits and costs remain unclear. The objective of the proposed research is to conduct an assessment of village-based community-based forest management in southern Malawi; more specifically to find out what "works," what does not, and why. CBNRM and the common-pool resource (CPR) literature that informs it, conflates the village to a homogeneous "community," and focuses on internal organizational matters that depend on sets of pre-defined "design principles," The literature ignores the roles of human agency, power relations, ecological agency, and scale in shaping human-environment interactions. The research uses a holistic political ecology approach that will combine remote sensing and GIS techniques, insights in landscape ecology; and spatial statistical models to supplement the social analysis of CBNRM. Main research hypotheses are that (1) the village-based CBNRM model does not fully achieve conservation objectives because of a partial mismatch between the ecological scale of forest fragmentation processes and social scales of forest conservation, and (2) the CBNRM model creates new institutions that reconfigure existing power relations and alter access to woodland resources in ways that can lead to forest fragmentation at the local scale.