Concerns over rapid widespread changes in social-ecological systems and their consequences for biodiversity, ecosystem functioning, food security, and human livelihoods are driving demands for globally comprehensive knowledge to support decision-making and policy development. Claims of regional or global knowledge about the patterns, causes, and significance of changes in social-ecological systems, or ‘generalized knowledge claims’ (GKCs), are generally produced by synthesis of evidence compiled from local and regional case study observations. GKCs now constitute a wide and varied body of research, yet they are also increasingly contested based on disagreements about their geographic, temporal, and/or thematic validity. There are no accepted guidelines for detecting biases or logical gaps between GKC’s and the evidence used to produce them. Here, we propose a typology of GKCs based on their evidence base and the process by which they are produced. The typology is structured by three dimensions: i) the prior state of knowledge about the phenomenon of interest; ii) the logic of generalization underlying the claim; and iii) the methodology for generalization. From this typology, we propose a standardized approach to assess the quality and commensurability of these dimensions for any given GKC, and their ability to produce robust and transparent knowledge based on constituent evidence. We then apply this approach to evaluate two contested GKCs – addressing global biodiversity and large-scale land acquisitions – and in doing so demonstrate a coherent approach to assessing and evaluating the scope and validity of GKCs. With this approach, GKCs can be produced and applied with greater transparency and accuracy, advancing the goal of actionable science on social-ecological systems.