The debate on improving access to medicines in developing countries has focused attention on the macroeconomic context and defining the demand side needs, based upon disease categories and purchasing power in a range of middle-income and Least Developed Countries. Far less attention has been given to understanding the diversity of supply side processes and their likely future evolution. This paper examines alternative frameworks for empirical analysis of supply side activities, namely, the manufacture and distribution of medicine, through the application of New Institutional Economics (NIE) concepts. Attention is focused particularly upon the potential utility of ideas from agency theory, transaction cost analysis and contemporary ideas from strategy theory. The major purpose of this paper is to use these theoretical frameworks to provide insight for policy makers, when faced with specific situations, whether in an international agency, or a private company, or in defining a national strategy. The analysis attempts to show the importance of distinctions between ideas of ‘make’ or ‘buy’, between ‘national self sufficiency’ and ‘international purchasing’ strategies, the limitations of contractual agreements under market governance and the crucial linkages between strategy formulation, strategy implementation and the necessary capabilities to achieve successful performance in practice. The current international situation on the investment, location and capacity of pharmaceutical manufacturing is reviewed and likely future scenarios suggested. Correspondingly current patterns of trade in medicines and their likely development within the context of the WTO and bilateral trade agreements are discussed. Against this background the promise and the pitfalls for new forms of public-private partnerships, which may offer attractive alternatives to conventional structures are evaluated. The implications of alternative future strategic options for national governments in setting the balance between health and industrial policies are examined and in particular the extent to which a national manufacturing capability should be developed or sustained. Similarly the scope for improving low cost distribution systems for medicines, based upon a mix of public and private sector channels, is assessed. We conclude with suggestions for further development of a transaction-based framework.