Once children and families become involved with the child welfare system, they are usually referred to a myriad of social services (e.g., concrete services, behavioral, mental, or physical health services). To examine a broad array of social services and child well-being over time, this study utilized secondary analysis of the National Survey of Child and Adolescent Well-Being via structural equation modeling with complex sampling techniques. Results showed that child well-being at baseline was significantly related to well-being over time. As social services were added to the model, child well-being decreased (i.e., significant partial mediation). There may be a trend to load higher-risk families with services. Trends like this suggest a need for more exploration of programming that directs resources toward lower-risk families to halt the cyclic nature of child welfare involvement. Practitioners and researchers are urged to consider the need for systematic measurement of service and dose. Close empirical examination will ultimately improve the effectiveness of the service delivery system and consequently the well-being of children and families engaged in child welfare services.