Multi-agency policing partnerships enable police agencies to address underlying social issues that contribute to crime and allow for longer-term solutions and potentially crime reductions. The success of these initiatives depends on how well police initiate and conduct these engagements. However, the concept of policing partnerships is not well understood and under-theorised. A better understanding of the process and system of engagement in these partnerships is needed. This is the central aim of this thesis: to identify and explore key factors and processes that facilitate or hinder multi-agency policing partnership implementation and engagement. In addition to identifying and exploring the key factors and processes in partnerships, the role of legal levers held by non-policing partners is informed by the third-party policing literature. The selection and engagement of non-police agencies and individuals because of the availability of legal levers to perform a crime prevention or control role underpins third-party policing theory. However, it is often assumed in this theory that police are aware, have knowledge and understanding of the legal levers available to partners and that the application of these legal levers is co-opted or coerced, by police. Whether these assumptions occur in such a clear manner has received limited attention in the literature. This is the second aim of the thesis: to gain an understanding of the role and use of legal levers in multi-agency policing partnerships, and particularly examine whether police conceptualise partnerships based on the availability of legal levers or partners, or whether legal levers come into play during formation, and to what extent they play a role in the maintenance of the partnership. To explore these two aims an international case study methodology was applied. Three case studies of partnerships addressing different crime and social issues across Australia and England were selected. The partnerships focused on addressing: 1) domestic and family violence in South Brisbane, Queensland, in the Community Coordinated Response (CCR) group; 2) alcohol-fuelled violence in North Brisbane, Queensland, in the Liquor Industry Accord Group (LIAG); and 3) minimising the risk of harm that serious violent and sexual offenders pose in England in the Multi-Agency Public Protection Arrangements (MAPPA). Using the collective case study method provided the opportunity to explore a third aim: to understand the perceptions and expectations of the police in multi-agency policing partnerships and to explore to what extent partnerships involving police vary across jurisdictions and partnership types. A total of 55 semi-structured interviews with police and partner agencies across 33 agencies, along with extensive documentary analysis were conducted.