The term "evidence-based medicine" first appeared in the medical literature in 1992 and is in widespread use today. It is timely to examine the concept's relevance to anaesthetic practice as well as the validity of the premises on which it rests. An important difference between anaesthesia and the specialties which treat disease is that in anaesthesia there is very little research done with real outcomes as its end-point. Surrogate or intermediate outcomes predominate as the end-points of anaesthetic research, which is a weakness when the results serve as the evidence on which to base clinical decisions. Furthermore, in interventions which require personal skill, dexterity or decision-making, caution must be exercised in assuming that equally good outcomes are achievable by all. Key members of the Cochrane Collaboration, among the most prominent advocates of evidence-based medicine, promote the belief that much valid scientific evidence is to be found in sources outside the peer-reviewed published literature. This assertion must be treated with caution. Furthermore, some techniques central to the search for the evidence on which to base practice, including meta-analysis and multicentre trials, are prone to errors through incorrect application. Evidence-based medicine appears to have less to offer anaesthesia than it does to the "treating" specialties.