This study investigates current teaching practice of the classroom religion program in two preschool settings in the Archdiocese of Brisbane. It also examines the approach to, and construction of, religious education in key Church and Brisbane Archdiocesan documents. Since the first Catholic school opened in Australia in 1820, research and scholarship have elucidated deeper understandings of the nature and purpose of religious education. Over time, a variety of approaches and curriculum models for the classroom religion program have been implemented in both primary and secondary schools. Broadly speaking there are two approaches to the Catholic primary school classroom religion program: educational and catechetical. The educational approach does not presume students' faith, and aims to develop students' religious literacy. The catechetical presumes student faith and aims to develop it. Currently, the Brisbane Catholic Education Religious Education Guidelines (Barry & Brennan, 1997a, 1997b; Barry et al., 2003) adopts an educational approach to the classroom religion program. However, while the approach to religious education in Catholic primary and secondary schools has received scholarly and professional attention over the years, the nature and purpose of religious education in early childhood education in the context of the Catholic preschool, have received minimal attention. Although the first preschools in Catholic schools in the Brisbane Archdiocese opened in 1988, there is no set curriculum for the classroom religion program for the preschool sector. However, Brisbane Catholic Education is presently preparing such a document in preparation for the introduction of the preparatory year of schooling into all Archdiocesan Catholic schools in 2007. The specific focus of the study is to use teachers' talk-in-interaction with their students during classroom religion lessons, as a means to exemplify their approaches to, and constructions of, their classroom religion programs. Underpinned by an Ethnomethodological methodology, the study gathered data in the form of lesson recordings from two preschool teachers. The lesson transcripts are analysed using the ethnomethodological analytic tools of Conversation Analysis and Membership Categorisation Analysis. These analyses reveal deep insights into teachers' practices: the nature of the content they present, their approaches to, and constructions of, their religion programs, as well as the ways in which they construct their students. In addition to classroom practice, this study also investigates relevant sections of the key Church documents The Religious Dimension of Education in a Catholic School (Congregation for Catholic Education, 1988) and the General Directory of Catechesis (Congregation for the Clergy, 1997), as well as the Archdiocese of Brisbane Catholic Education document, 'Religious Education in Preschools', which is part of the Preschool Handbook: Towards Continuity of Learning in the Early Years (Catholic Education Archdiocese of Brisbane, 2002b). Together with the Ethnomethodological methodology, this part of the investigation adopts a functional linguistic methodology using the analytic technique, Systemic Functional Linguistics. Both Systemic Functional Linguistics and Membership Categorisation Analysis are used to explicate these documents. The two Church documents are critical documents, as they contribute to curriculum development and implementation of the classroom religion program in all Australian Catholic schools, whilst the Brisbane document outlines the current policy for religious education in Catholic preschools in the Archdiocese. These analyses elucidate key insights into how the classroom religion program is approached, and reveal that whilst the Church documents maintain an educational approach, aspects of the documents are ambiguous. Analysis of the Brisbane Archdiocesan preschool document reveal it to be at variance with the current educational approach taken by the Archdiocese in its classroom religion curriculum for primary and secondary schools. This study contributes significantly to the nature and purpose of religious education in the early years. It has implications for the theory and practice of the classroom religion program in early childhood, and for preservice and inservice teacher education programs. It also contributes to policy design that guides and shapes curriculum development and implementation. The use of analytic techniques drawn from two different methodologies, Ethnomethodology and functional linguistics, enables a detailed and in-depth analysis, showing them to be effective techniques to be used together in research. These methodologies complement each other to reveal critical insights into both the document studies and teacher classroom interaction. The nature and purpose of religious education in early childhood education is evolving. As Catholic dioceses continue to expand into early childhood education, the focus on religious education in this sector becomes more critical. This study provides a significant foundation for future research.