Scenarios have become a standard tool in the portfolio of techniques that scientists and policy-makers use to envision and plan for the future. Defined as plausible, challenging and relevant stories about how the future might unfold that integrate quantitative models with qualitative assessments of social and political trends, scenarios are a central component in assessment processes for a range of global issues, including climate change, biodiversity, agriculture, and energy. Yet, despite their prevalence, systematic analysis of scenarios is in its beginning stages. Fundamental questions remain about both the epistemology and scientific credibility of scenarios and their roles in policymaking and social change. Answers to these questions have the potential to determine the future of scenario analyses. Is scenario analysis moving in the direction of earth system governance informed by global scenarios generated through increasingly complex and comprehensive models integrating socio-economic and earth systems? Or will global environmental scenario analyses lose favour compared to more focused, policy-driven, regionally specific modelling? These questions come at an important time for the climate change issue, given that the scenario community, catalyzed by the Intergovernmental Panel on Climate Change (IPCC), is currently preparing to embark on a new round of scenario development processes aimed at coordinating research and assessment, and informing policy, over the next five to ten years. These and related questions about where next to go with global environmental scenarios animated a workshop held at Brown University (Note1) that brought together leading practitioners and scholars of global environmental change scenarios from research, policy-making, advocacy, and business settings. The workshop aimed to provide an overview of current practices/best practices in scenario production and scenario use across a range of global environmental change arenas. Participants worked to bring the experience generated from over four decades of scenario development in other issue domains, including energy and security, to bear on environmental scenarios, and to bring into dialogue scenario practitioners, both producers and users, with social science scholars. The set of contributions to this focus issue of Environmental Research Letters arose out of this workshop and collectively examines key challenges facing the scenario community, synthesizes lessons, and offers recommendations for new research and practice in this field. One theme that emerged in many of the discussions at the workshop revolved around the distinction between two broad perspectives on the goals of scenario exercises: scenarios as products and scenarios as processes. Most global environmental change scenario exercises are product-oriented; the content of the scenarios developed is the main goal of many participants and those who commission or organize the scenario development process. Typically, what is of most interest are the environmental outcomes produced, how they relate to the various factors driving them, and what the results tell us about the prospects for future environmental change, for impacts, and for mitigation. A product-oriented perspective assumes that once produced, scenario products have lives of their own, divorced from the processes that generated them and able to serve multiple, often unspecified purposes. Thus, it is often assumed that the scenario products can be 'taken up' by a variety of users in a variety of fora. A contrasting scenario approach is process-oriented and self-consciously privileges the process of scenario development as the primary goal, for example as a means to motivate organizational learning, find commonalities across different perspectives, achieve consensus on goals, or come to a shared understanding of challenges. Focusing on scenarios as processes highlights the social contexts in which scenarios are created and used. Process-oriented scenario exercises also generate scenario products, but such products are recognized as meaningful mostly (or only) in the social context in which they were developed. It should be noted that those seeking to understand the functions, implications and utility of scenarios can approach analysis of scenarios and their impacts from either perspective—focusing attention on product outcomes and influence or assessing procedural and contextual dynamics and implications. Papers in this issue examine various aspects of scenario products, scenario processes and their interactions, with specific reference to global environmental change scenarios. Hulme and Dessai (2008) use the product–process distinction as a starting point for developing a framework to evaluate the success of scenario exercises. They identify 'prediction success', 'decision success' and 'learning success' as three evaluation metrics for scenarios, with the first two most relevant to scenario products and the last emphasizing procedural aspects of scenarios. They suggest that viewing scenarios primarily as products implies examining how closely actual outcomes have matched envisioned outcomes, while viewing them primarily as processes suggests evaluating the extent to which scenarios engaged participants and enabled their learning. O'Neill and Nakicenovic (2008) focus on Hulme and Dessai's evaluation metric, learning. Based on a review of six scenario/assessment exercises, they ask if and how scenario products have incorporated comparative assessments of results in order to enable cumulative learning across scenario efforts. The authors conclude that, although participating modelling teams have benefited greatly from the process of scenario activities and applied that learning to other scenario exercises in which they engage, learning from comparative assessments of scenario products has been rather limited; the latter due to the limited time and resources...