This thesis seeks to formulate a theory that explains the geographically uneven development of privatisation. The New Zealand public hospital ancillary services are taken as a case study of uneven privatisation, and the process of privatisation is analysed within terms of the three major theoretical frameworks used in social geography; public choice theory, Weberian managerialism and Marxian class conflict. Empirically it is found that geographically uneven privatisation occurs within three dimensions; the spatial, the temporal and the sectoral. This raises the questions, why privatisation in region A but not B, why at certain times and why in certain industries or sectors of industry but not in others? The answers to these questions are sought in the context of the three theoretical frameworks used in the thesis, as applied to hospital ancillary services. The empirical research reveals that the implementation of privatisation policies is mediated by several regionally variable factors, such as the size of institutions, hospital authority management structures, location to major urban settlements, levels of financial restraint on hospitals, and the labour militancy of the work force. The explanatory significance of these factors varies substantially according to the time and place at which privatisation occurs and the particular sector of the public hospital system being considered. Many of the explanatory factors contradict each other, insofar as some may predispose public authorities to privatise whereas others may inhibit the process. The complexity of the uneven development of privatisation renders inadequate any single theoretical explanation of the process. Nevertheless the research presents an understanding of how privatisation develops across space and time and an insight is also gained into the likely future uneven development of public service provision.