INTRODUCTION: The ideology and pronouncements of the Australian Government in introducing 'competitive neutrality' to the public sector has improved efficiency and resource usage. In the health sector, the Human Services Department directed that non-clinical and clinical areas be market tested through benchmarking services against the private sector, with the possibility of outsourcing. These services included car parking, computing, laundry, engineering, cleaning, catering, medical imaging (radiology), pathology, pharmacy, allied health and general practice. Managers, when they choose between outsourcing, and internal servicing and production, would thus ideally base their decision on economic principles. Williamson's transaction cost theory studies the governance mechanisms that can be used to achieve economic efficiency and proposes that the optimal organisation structure is that which minimises transaction costs or the costs of exchange. Williamson proposes that four variables will affect such costs, namely: (i) frequency of exchange; (ii) asset specificity; (iii) environmental uncertainty; and (iv) threat of opportunism. This paper provides evidence from a rural public hospital and examines whether Williamson's transaction cost theory is applicable. METHOD: Case study research operates within the interpretivism paradigm and is used in this research to uncover why the outsourcing decision was made. Such research aims to study real-life experiences by examining the way people think and act and, in contrast to positivism, allows the interviewer to participate to better understand the details and features of the experiences. In the present research, individual interviews were conducted with managers of the hospital and owners and staff of the vendor organisations using semi- and unstructured questions to ascertain the extent of, and processes used in, outsourcing specific functional areas, and areas that were not outsourced. RESULTS: Pathology, radiology, dental technician services and lawn mowing were outsourced while food services was retained internally. The outsourcing of radiology was due to the hospital being unable (or unwilling) to finance new equipment and the problematical relationship between the existing radiologists, and hospital management and staff. Outsourcing resulted in increased staff morale, upgraded capital equipment and improved services. The outsourcing of pathology and dental technical services aimed to increase labour flexibility, thereby decreasing costs. Additional drivers in pathology were the changing nature of the funding arrangements rendering it profitable for the private sector to move into the provision of pathology and the increasing power of the medical scientists' union. The outsourcing of lawn mowing was simply to reduce costs. Food services was not outsourced because there was a lack of evidence that costs could be reduced. In addition, the existing relationships with food services staff were regarded as important because they had previously made immense changes to work practices, reduced staff numbers and decreased costs. CONCLUSION: Transaction costs are important when analysing how managers make the outsourcing decision, but the evidence from this case is that not all transaction costs are included in the decision, and that such costs are more complex than can be included in the type of analysis often undertaken by decision-makers. Taking into account Williamson's variables, the research shows that the outsourcing of services did not comply solely with the levels of transaction frequency or the requirement of asset specificity. In addition, opportunistic behaviour was evident on the part of all parties and was used in some cases as a reason for outsourcing, and in others to sway the decision to the manager's predisposed choice. A variety of arrangements were used to reduce environmental uncertainty, such as the transfer of staff to the contractor and the use of long-term contracts. Indeed the case shows that relationships between the hospital, its staff and the vendor are an important consideration that may not always be factored into an analysis that relies solely on transaction costs.