Purpose: The casework relationship is viewed as a critical practice component and key change factor in many social work interventions. Yet, many evaluations do not focus on the process by which desired outcomes are achieved. It is important for knowledge development in our field that we break down and evaluate the pathways from theories of change to anticipated effects. Despite considerable efforts to implement family preservation programs in which the worker-client relationship is a central feature of services to clients, evaluations of these programs have not shown them to be more effective than usual child welfare services in preventing foster care placements, reducing rates of subsequent maltreatment, or improving child or family functioning . Understanding the role of the worker-client relationship and the context in which such relationships are formed will contribute to researchers' and practitioners' understanding of a critical aspect of service delivery in child welfare and other social work interventions. Method: In this study, data from the Evaluation of Family Preservation and Reunification Programs were used to assess caregivers' reports of the quality of their relationships with caseworkers. This evaluation was conducted in four states and families were randomly assigned to receive either regular child welfare services or family preservation services. Data were drawn from interviews with the caseworkers and caregivers, surveys of caseworkers, and administrative records. Ordinary least squares (OLS) regression, hierarchical linear modeling (HLM), and logistic regression were used to a) identify client, service, and caseworker factors associated with the quality of the relationship, and b) examine whether the quality of the worker-client relationship was associated with client outcomes. Results: Overall, reports of the quality of the relationships were positively skewed, suggesting that caregivers were generally pleased with their caseworkers. However, an assessment of the variance, mean, and median values for the worker-client relationship revealed considerable variability in relationship quality both within and among the four different states and two types of programs. Clients' race or ethnicity and report of depressive symptoms, caseworkers' job satisfaction and education, and the level of worker activity and contact were all important factors associated with the quality of worker-client relationships. Furthermore, the quality of the worker-client relationship was associated with clients' sense of improvement and positive parenting practices. Implications for Practice: This study supports an ongoing emphasis on the role of good quality worker-client relationships in the delivery of child welfare services. Study results point to the significance of the allocation of workers' time and resources as well as the importance of the context of casework. The findings also have implications for the implementation of family preservation programs as well as caseworker training and the monitoring of service delivery – all with the goal of motivating and supporting clients.