Although often neglected, model validation is a key topic in flood risk analysis, as flood risk estimates are characterised by significant levels of uncertainty. In this paper, we discuss the state of art of flood risk models validation, as concluded from the discussion among more than 50 experts at two main scientific events. The events aimed at identifying policy and research recommendations towards promoting more common practice of validation, and an improvement of flood risk models reliability. We pay specific attention to the different components of the risk modelling chain (i.e. flood hazard, defence failure and flood damage analysis) as well as to their role into risk estimates, to highlight specificities and commonalities with respect to implemented techniques and research needs. The main conclusions from this review can be summarised as the need of higher quality data to perform validation and of benchmark solutions to be followed in different contexts, along with a greater involvement of end-users in the debate on flood risk models validation.