Summary This interview with Brian Walker, chair of the research-based Resilience Alliance, outlines the main concepts and propositions behind ‘resilience thinking’ and touches on the importance of this paradigm for individuals and organizations involved in managing complex social-ecological systems. It refers to the origins, work and publications of the Resilience Alliance, listing and elaborating the key case studies used to illustrate the Alliance's main proposition that complex social-ecological systems do not behave in a predictable linear fashion. Rather, research indicates it is normal for complex systems to go through cycles of increasing and decreasing resilience and to have potential to shift, (in a self-organising way) to potentially undesirable states or entirely new systems if certain component variables are severely impacted by management. Such shifts can be novel and ‘surprising’, and are often not beneficial or desirable for societies. This is particularly the case where small-scale solutions push the problem upwards in a system, causing loss of resilience at a global scale. Predicting thresholds is therefore important to managers and is a key research focus for members of the Resilience Alliance who are currently building an accessible database to support decision-making in global natural resource management.