This paper presents a methodology to help countries identify and characterise global emerging critical risks as part of the OECD’s Framework on the Management of Emerging Critical Risks. It supports experts and policymakers tasked with anticipating and preparing for uncertain and evolving threats that transcend traditional national boundaries. The approach begins with horizon scanning to capture weak signals and unconventional data sources, including patent analysis, crowd forecasting, and the use of generative AI. It then applies structured foresight techniques, such as futures wheels, cross-impact analysis, and scenario-based “Risk-Worlds,” to explore how risks might manifest and interact in multiple possible future contexts. The methodology emphasises understanding risks “at source,” focusing on vulnerabilities, interconnectedness, and possible management strategies. Rather than predicting a single future, it seeks to broaden the range of possibilities, encouraging proactive adaptation, building collective understanding, and ultimately strengthening government capacity to navigate and shape an increasingly complex and uncertain global risk landscape.