Health care now accounts for just under 10% of national income in most developed OEeD countries. Yet, despite its importance in the overall economy, the factors that drive health care expenditure remain only imperfectly understood. This thesis shows that the determinants of health care expenditures are simply too diverse amongst different countries to be brought within a common denominator such as GDP, as it has been argued for decades. It also shows that the assumption that health care is a homogeneous good across countries is over-simplistic and arbitrary, and finds ample evidence showing that health care is not a luxury good, as widely suggested. The contribution of the thesis is on methodological, theoretical and empirical grounds. In terms of methodology, the thesis shows that there are significant flaws in several areas that influence our thinking concerning the determinants of health care expenditures and offers alternative ways of analysis and appraisal. Flaws were shown in: the relationship between health expenditure and GDP; the importance of factors such as ageing; the macroeconomic context and the burden of disease; the measurement of key variables used in empirical analysis such as health spending, national income, technology, and health prices; the method of analysis that has been pursued; and the conversion factors used to translate prices and monetary variables across countries into a single and comparable denominator. The thesis makes a theoretical contribution of the analysis of health care expenditures, assuming that health care is at least a quasi-public good. The proposed conceptual framework explicitly links the determinants of health care expenditures to the theory of public finance and allows flexible adjustments by decision-makers to account for changes in technology, prices, and the macroeconomic environment. The impact of the macroeconomy on health spending is assessed by evaluating whether the rate of growth of income has any influence on the demand for health and whether the fiscal deficit impacts on health spending and to what extent. The proposed framework incorporates technology and this is an advance from the published literature, which has almost invariably considered technology to be a residual factor. Finally, the thesis recognises that the lag structure of the model, the availability of data, and knowledge of the relationship between disease and need for services are not sufficient to test for the impact of lifestyle and disease factors on health spending. The empirical investigation provides conclusive evidence of the non-importance of GOP in explaining health care spending trends over time. Consumption is shown to be a predictor of health expenditures; technology is an important cost-push factor across countries; the macroeconomy exerts, in general, significant pressure on health care expenditure; however, the impact of health care reforms does not show any significant impact on health care expenditures; and the number of doctors per capita has little or no association with health care expenditures.