THIS LITERATURE REVIEW COVERS journal articles reporting early childhood education (ECE) research in Australia and New Zealand in the past 10 years and it has found that: (1) Nearly half of the articles lacked child-related data with an explicit or implied justification, while less than one-tenth failed to include child-related data with no justification; (2) Over one-third of the articles addressed the voice of the child, while less than one-tenth included child-related data for assessment purpose. A typology of the voice of the child has evolved from the review and it includes: pseudo voice (engendered by assessing the child), inferred voice (engendered by inferring the child's perspective), surveyed voice (engendered by surveying the child) and co-constructed voice (engendered by co-constructing lived experience and understanding with the child). In spite of the evidence from the articles that the voice of the child had generally been recognised, the review has identified a gap between the child's voice rhetoric and research practice. It is argued that the distinction between rhetoric and practice should be recognised and respected.