One aim of the Evidence-based Practice in Science Education (EPSE) Network was to obtain a better understanding of the extent to which practitioners in science education recognise and make use of research findings in the course of their normal practice. The aim was realised through an interview and focus-group study of views of practitioners on the research-practice interface. The sample included primary and secondary science teachers (including a subset with direct experience of research), curriculum policy-makers, in-service trainers and authors of science textbooks and teaching materials. Questions explored perceptions of the nature, actual use and potential of research in science education. At a general level, teachers and other practitioners characterised educational research, variously, as: purposeful; carried out in a systematic manner; useful in informing action; and large scale. To be viewed as convincing, research evidence had to be seen as transferable, to have resonance with teachers experience and beliefs, and to have a rigorous methodology. Widespread use of research evidence in the classroom seems to depend on at least two factors: translation of research findings into tangible and useful outcomes, such as teaching materials resulting from research projects; and the presence of a professional culture which encourages both exploration of research and changes to practice. Increasing use of research evidence requires researchers to translate outcomes of research into practical actions and practitioners to be skilled in using evidence systematically in evaluation of their own practice. Networks of researchers and practitioners may further the aims of evidence-based practice.