Further methods for Stochastic Impact Evaluation and apply them to improving stakeholder intervention decisions in WLE priority regions. This is a cross-cutting initiative that addresses how the CGIAR’s stakeholders, and the CGIAR itself, can improve how they prioritise, plan, monitor, and evaluate projects and interventions through quantitative decision modelling under uncertainty. The initiative aims to further develop methods, applications, and capacity that will help CGIAR stakeholders to (i) quantify the impact pathways, costs, benefits, and risks associated with planned interventions, (ii) identify high priority metrics for monitoring during implementation, and (iii) accumulate evidence for impact of the intervention along the impact pathway. This will in turn help prioritize which interventions would most impact on the System Level Outcomes at least cost and risk, and accelerate reliable learning on which factors lead to or constrain success. The overall aim of this initiative is to mainstream use of decision analysis under uncertainty in agricultural research and development. The objectives are to: 1. Develop tools and guidelines for Stochastic Impact Evaluation in agricultural development, including building a probability management system. 2. Analyse a sample of real decision cases in partnership with key stakeholders across the CGIAR portfolio to provide proof of concept and examples that can be applied to new decision cases. This includes modelling of impact pathways at CGIAR, CRP and Centre levels. 3. Develop a cohort of stakeholders, including in the CGIAR, with capacity for decision analysis under uncertainty. 4. Roll out a dissemination and capacity building programme to achieve wide-scale adoption of the key principles, approach and methods.