High quality scientific evidence is increasingly available to help the providers of health care make decisions about what types of investigations to order for\nan individual patient, and how to treat any health problems that are identified. “Evidence-based medicine” has been facilitated by the availability of results\nfrom different types of clinical trials, epidemiological studies, and meta-analyses which allow the results of many small studies to be combined in a way that reduces the uncertainty about the overall effect of clinical interventions.\n\nThis evidence is critical to the way that the World Health Organization fulfils part of its mandate, the part linked to setting standards and providing technical advice on clinical issues. To illustrate, WHO recently developed a set of internal “guidelines for guidelines” that specify the steps that should be taken to ensure any clinical guidelines released or endorsed by WHO are based on the best available scientific evidence (URL: http://www.who.int/health-systems- performance/).\n\nWhile clinical guidance is one important part of WHO’s mandate, the Organization also receives continual requests from countries to provide advice on how best to organize and manage health systems. It is critical that this advice be based on rigorous scientific evidence, but when I took office in 1998, it became clear that there was little systematic evidence on what makes health systems perform well and what makes them perform badly. One reason was that there were few clear statements of what health systems were supposed to achieve, so the case-studies that existed measured outcomes and goal attainment in a variety of ways. Another was that it is clearly much more difficult to base evidence in this area on the same types of trials that are typically used to show evidence of clinical effectiveness.\n\nIt is rarely possible, for example, to randomize parts of a country into experimental groups to assess the impact of different types of health system reforms. Policy advice on health system development has, until recently, been based on case-studies and, some- times, ideology. Case-studies can be useful, just as case-\nreports provide valuable information to clinicians. On the other hand, evidence-based clinical medicine can- not exist without epidemiological studies, which operate on the premise that, although all individuals are different in many ways, there is much to be learned from what they have in common. This is also important in the area of health systems performance. All systems and cultures are different in many ways, but there is a great deal of knowledge to be gained from\nthe experiences of groups of countries taken together, learning from common experience.\n\nThis volume reports on five years of work to strengthen the scientific evidence-base on health sys- tems performance. It began with the development of a framework that clearly specifies a parsimonious set of key goals to which health systems contribute, a framework widely discussed with experts, policy-makers and the governing bodies of WHO. After a series of consultations on specific components, the first set of figures on goal attainment and health system efficiency was published in The World Health Report 2000 for all 191 countries that were then Members of the Organization. That report generated an enormous amount of\ninterest and debate among policy-makers and the scientific community, and it was decided that WHO should report on the performance of the health systems of its Member States at regular intervals. To facilitate this, a further series of consultations was held, and I established an independent Scientific Peer Review Group to review the techniques proposed by the Secretariat for future rounds of performance assessment.\n\nI am delighted to introduce this volume, which traces the history of this work, openly reports the debates and criticisms, and describes the methods proposed\nfor future rounds of performance assessment after the peer review and consultation process. This is not the end of the work. Science advances over time with open debate. I am sure the techniques and methods will develop further over the next decades, but I am also sure that the area of health systems performance assessment is now moving in the right direction. The movement towards basing health policy development and advice on rigorous scientific evidence has begun.