Anthropogenic changes to the Earth’s climate are increasing the frequency and intensity of natural hazard events, creating intractable environmental changes, challenging scientific assumptions and, thereby, prompting new responses to the governance of different biomes. If, to a certain extent, environmental forces and entities such as weather and forests need to be stabilised as objects of knowledge in order to become sites of action or intervention, then this situation presents both new and old epistemic challenges for those who predict and those who seek to conceptualise prediction. This session brings together a range of contributions focused on sites and scenes of predicting, forecasting, or providing foresight that stage ecologies of prediction in fluid and uncertain worlds. This phrase - ‘ecologies of prediction’ - is a pun on philosopher of science Isabelle Stenger’s propositional call for an ecology of practices, which itself points to the heterogeneous ways in which knowledge practices both describe and make “what matters to them matter” in contexts of degradation, destruction or change (Stengers 2015, 112). Conceptually, thinking with ecologies of prediction means retaining this sense of heterogeneity and urgency while recognizing that predictive knowledge claims emerge and circulate within various ‘ecologies’ • epistemic, social, political, and material formations of agents. This session explores how these predictive ecologies take hold and will ask: how do they make anthropologists, publics and others think differently about environmental threats? What sort of commitments allow them to ‘hit their mark’ or diverge from adverse or destructive events in fluid situations? Without resort to general rules of ‘truth’, and along with Stengers, we approach these ecologies both as they are, “as they may become” in relation to climatic and social changes. References: Stengers, Isabelle. 2015. In Catastrophic Times: Resisting the Coming Barbarism: Open Humanities Press.