This thesis explores the connection between policy creation and social work practices that are related to this policy. The main aim of the thesis is to fill a gap that exists in the research in relation to the connection between particular policies concerning the care and protection of children and the social work practices related to these. Primarily, I am interested in the experiences of social workers in community groups and the issues and problems they face in trying to integrate these policies into their everyday practice. The thesis presents four case studies that highlight the interactive relationship that exists between policy and practice. These case studies have been developed from the interviews I conducted with care and protection community workers in Christchurch in 1999. The first two case studies - the development of the Children, Young Persons and Their Families Act (1989), and the process of devolution that occurred through the 1980s and 1990s - look at particular policy developments that have impacted on care and protection social work, and explore the way that policy creation and implementation is contingent on the specific time and place in which it is developed. The second two case studies - Family Group Conferences and Strengthening Families - focus on two quite different forms of social work practice and provide an insight into the way that policy is implemented and used by practitioners at ground level. These four case studies form the basis of an argument around the idea that policy and practice are dynamic and interactive processes that will inform and change one another. Rather than seeing policy as something that is created by bureaucrats in the state and applied by practitioners at the ground level, I argue that the policy process is more complex than this. The case studies provide practical examples of this idea, and explore the complexities of policy development and the relationship between policies, policy actors and specific community social work practices - an area about which there has been little research.