The Sendai Framework for disaster risk reduction (SFDRR) and its seventh global target recognizes that increased efforts are required to develop risk-informed and impact-based multi-hazard early warning systems. Despite significant advances in disaster forecasting and warning technology, it remains challenging to produce useful forecasts and warnings that are understood and used to trigger early actions. Overcoming these challenges requires understanding of the reliability of forecast tools and implementation barriers in combination with the development of new risk-informed processes. It also requires a commitment to create and share risk and impact data and to co-produce impact-based forecasting models and services. To deal with the problem of coming into action in response to imperfect forecasts, novel science-based concepts have recently emerged. As an example, Forecast-based Financing and Impact-based Multi-Hazard Early Warning Systems are currently being implemented operationally by both governmental and non-governmental organisations in several countries as a result of increasing international effort by several organizations such as the WMO, World Bank, IFRC and UNDRR to reduce disaster losses and ensuring reaching the objectives of SFDRR. This session aims to showcase lessons learnt and best practices on impact-based multi-hazards early warning system from the perspective of both the knowledge producers and users. It presents novel methods to translate forecast of various climate-related and geohazards into an impact-based forecast. The session addresses the role of humanitarian agencies, scientists and communities at risk in creating standard operating procedures for economically feasible actions and reflects on the influence of forecast uncertainty across different time scales in decision-making. Moreover, it provides an overview of state-of-the-art methods, such as using Artificial Intelligence, big data and space applications, and presents innovative ways of addressing the difficulties in implementing forecast-based actions. We invite submissions on the development and use of operational impact-based forecast systems for early action; developing cost-efficient portfolios of early actions for climate/geo-related impact preparedness such as cash-transfer for droughts, weather-based insurance for floods; assessments on the types and costs of possible forecast-based disaster risk management actions; practical applications of impact forecasts.