Health outcomes are becoming the currency of health care exchange, and a call for evidence dominates decision making at all levels. This discussion paper reviews methodological and sociopolitical barriers that impede the production and dissemination of outcome research in public health, with particular reference to nursing. Barriers to the production of high-quality research evidence include inaccessible graduate education and inadequate research funding. Also, randomized controlled trials (the ideal design for interventions studies) are uniquely difficult to implement for public health services. Practical and ethical difficulties arise in defining the intervention, implementing random allocation methods, selecting and measuring outcomes, and articulating adequate theoretical frameworks. When health care activity is defined as output, there is a tendency to exclude the ethical standing of preventive, supportive and communitarian functions. The production and interpretation of research results must remain part of a social, political and ethical debate, not a purely scientific one.