The purpose of the paper is to reflect on the recent health care reforms in both developed and developing countries, in the light of the evidence that has accumulated over the last few years about the efficiency and equity of different fiscal and organisational arrangements. The scene is set by a brief review of the definitions of efficiency and equity and of the confusions that often arise; and of the problems of making assessments in practice with real data. The evidence about effectiveness, efficiency and equity at the macro level are reviewed: among OECD countries, there is little evidence that variations in the levels and composition of health service expenditure actually affect levels of health; equity in financing and delivery appears to mirror equity in other sectors in the same countries; about the only solid--although rather limp--conclusion which is transferable is that costs can be contained best via global budgeting. The range of reforms in the North is sketched: despite calls to give people 'freedom' to opt out, public finances continues to be preferred among OECD countries; and the evidence that health care markets can actually function is 'weak'. Whilst geographical redistribution of finance has proved to be possible, inequalities in health remain in most countries. But the overwhelming impression is that the quality of the data base for many of these studies is appalling, and the analytice techniques used are simplistic. The move to introduce user charges in the South is discussed. It seems unlikely that they will raise a significant fraction of overall revenue; exemptions intended for the poor do not always work; and other trends are likely to exacerbate the patchy coverage of health care systems in the South. The final section reflects on the pressures for increased accountability. The emphasis on consumerism in the North has led to an increasing number of poorly designed 'patient satisfaction' surveys; in the South, there has been an increasing rhetoric on community participation, but little sign of actual devolution of control. The flavour of the decade is 'outcome measurement' which has been promoted feverish but with little rigour. We must also be concerned that this emphasis will, once again, be hijacked by the most articulate.