This article presents a framework for analysing the impacts of evidence-informed policy entrepreneurship, and applies it to a longitudinal policy research project on childhood poverty in the developing world. Drawing on insights from Keck and Sikkink's (1998) work on transnational advocacy, the discussion highlights five key dimensions of possible policy impact: framing debates and policy agenda formulation; securing discursive commitments from key policy actors; bringing about procedural changes; policy reform; and behavioural change. The analysis pays particular attention to the interplay of culture, politics and values in the uptake of research evidence in the policy process, and the challenges of cross-institutional partnership approaches to knowledge translation initiatives.