Multi-agency working (MAW) has increasingly been considered a promising approach to enable the early and effective identification of individuals and communities that are at risk of radicalisation and violent extremism. Multi-agency responses usually involve collaboration between local organisations and are based in the belief that the complex problems of radicalisation and violent extremism cannot be effectively addressed by one single agency. However, more than a decade after the conclusion that evaluation in the field of countering violent extremism (CVE) is still in its infancy, it remains underdeveloped and evaluations remain scarce. Challenges such as the lack of established MAW policies and procedures, and information-sharing barriers, have been reported in building effective MAW. The ‘Evaluation and Mentoring of the Multi-Agency approach to violent radicalisation’ (EMMA) project was established firstly to evaluate the MAW approach, and secondly to mentor peer-to-peer assessment and exchange best practice among local practitioners. It asked the question ‘What works under what conditions?’, assessing the approaches used in different countries by means of a realist process evaluation. This book reports the indicators of good MAW practices from a wide range of situations, and gives concrete recommendations for both practitioners and policy-makers. The EMMA project also resulted in the development of a website-based self-evaluation tool for use by local MAW practitioners that will be widely applicable across different MAW approaches in Europe.