Effective ecological monitoring is imperative in a human-dominated world, as our ability to manage functioning ecosystems will depend on understanding biodiversity responses to anthropogenic impacts. Yet, most monitoring efforts have either been narrowly focused on particular sites, species and stressors - thus inadequately considering the cumulative effects of multiple, interacting impacts at scales of management relevance - or too unfocused to provide specific guidance. We propose a cumulative effects monitoring framework that integrates multi-scaled surveillance of trends in biodiversity and land cover with targeted evaluation of hypothesized drivers of change. The framework is grounded in a flexible conceptual model and uses monitoring to generate and test empirical models that relate the status of diverse taxonomic groups to the nature and extent of human "footprint" and other landscape attributes. An adaptive cycle of standardized sampling, model development, and model evaluation provides a means to learn about the system and guide management. Additional benefits of the framework include standardized data on status and trend for a wide variety of biodiversity elements, spatially explicit models for regional planning and scenario evaluation, and identification of knowledge gaps for complementary research. We describe efforts to implement the framework in Alberta, Canada, through the Alberta Biodiversity Monitoring Institute, and identify key challenges to be addressed.