The increasing use of performance measurement by government means that child protection services are under pressure to demonstrate effectiveness in protecting children from harm and efficiency in the use of public funds to help children and families. From a policy perspective, the way performance measurement is conceptualised and implemented can have major consequences for service delivery. This paper examines key issues raised in the literature about performance measurement, the context for its introduction in child protection, how the concepts of effectiveness and efficiency are dealt with, how client outcomes are defined, and assumptions about 'good performance'. An overview of performance measurement in child protection in Australia is provided. The paper argues that a critical approach to performance measurement in child protection can contribute to improved service delivery to clients.