In this issue, M. James, Pollard, Rees and Taylor consider what research must do in order that 'practitioners and policy-makers will want to feel confident that the adoption of new research-based practices will repay their change efforts'. In responding to this collection of papers, it is suggested that users are a wider and more complex group than is generally assumed; that the research?user relationship needs to be reciprocal and multifaceted; and that considerable value can be drawn by users from research into learning outcomes in a number of ways over different time-scales.