In international development, impact evaluation (IE) is becoming more and more an institutionalised practice. This article starts out by addressing the question of what institutionalisation of IE means and how it could work. Subsequently, the article explores common challenges in monitoring and evaluation functions in the UN system related to the supply of (and to a lesser extent demand for) evidence on impact. Rather than looking for solutions to these challenges in the practice of IE, the article explores the issue of how to improve non‐IE monitoring and evaluation practices. On the basis of the identified challenges three categories of solutions are discussed: improving the quality of impact‐related evidence at activity and project level, strengthening the causal logic underlying interventions, and strengthening the aggregation and synthesis of evidence. Finally, the article presents some illustrative examples of the latter two categories of solutions.