This research examines the relationship between primary postgraduate (PGCE) students’ conceptions of geography, knowledge and pedagogy and their development as teachers of primary geography over two years – the primary PGCE course and the first year of teaching. The methodology is essentially qualitative and based on the principles of grounded theory. Personal Construct Theory (PCT) has also informed the choice of research techniques, PCT being seen to be appropriate for a research project that aims to access and therefore examine a range of alternative constructions. Concept mapping was used at the beginning and end of the geography component of a PGCE Primary course to elicit all students’ conceptions of geography, teaching and learning. Analysis of the concept maps from the whole cohort (n=79) show that primary students’ conceptions of geography are generally rather simplistic and reflective of the descriptive-rich and scientific persuasions identified by Barratt Hacking (1996). Only a few students’ maps reflected environmental or humanistic/welfare persuasions. The concepts maps were also sorted into four categories from most sophisticated to least sophisticated conceptions of geography. It was noted that of the eight students in category one (most sophisticated) only one had a geography degree. A sample of 11 students was then interviewed (using a stimulated recall technique (Calderhead 1986)) about their conceptions using the elicitation data as a stimulus for the discussion. This enabled the researcher to both probe students’ conceptions in greater depth, and to validate initial analysis of the elicitation data. Finally, three students – one a geographer (with a geography degree) and the others non-geographers – were observed teaching geography and interviewed directly after the observations on three occasions over the two years. A coding system was developed from all the data, and was then used to analyse the interviews using Microsoft Word index and cross-referencing functions. These analyses, along with elicitation data, formed the basis of case studies of the development of three students as geography teachers over two years. A model for beginning teacher development in the field of primary geography is then proposed.  The model emerged from interpreting and synthesising the evidence from the three case studies, and through the use of the constant comparison technique (Strauss & Corbin, 1998). The model is applied to the series of lessons observed for each case study providing an overview of their development as teachers of primary geography. Comparison of the three cases over two years shows some startling similarities as well as some differences in their development. It seems that each of them, whether they held a geography degree or not, discounted the geographical knowledge they have gained from life experiences as a valuable base to work from, despite the relevance of this knowledge to their teaching. It also seems that each of them, when observed during their PGCE course, were most likely to draw on their memory of geography lessons from when they were a pupil as a model to inform their teaching. As the two years progressed, and their pedagogical knowledge developed, they began to replace these early experiences with ones more suited to effective teaching – that is, their more recent experiences as teachers.  Of the three beginning teachers, only David, who had a geography degree, developed to become an effective geography teacher during the research period. However, it is considered that, for the majority of primary teachers, the most that can be expected is that they will develop into effective teachers of primary geography because it is unlikely that they would have opportunities to develop the depth of subject knowledge required to be an effective geography teacher. The thesis concludes by offering some thoughts for the development of primary geographical education. It proposes that primary geography could be usefully conceptualised as ‘everyday’, or ‘ethno-‘geography, that is a geography that recognises and seeks to address the ‘false split between practical, everyday knowledge and abstract, theoretical knowledge’ (Frankenstein & Powell, 1994). This is a geography that explicitly values the geographical knowledge that we all build up from everyday experiences in the world and that, in conjunction with the development of a geographical imagination, might form the basis of a primary geography framework.