In ever larger areas of the world, climate change is increasing water demand while shrinking water supplies. As a result, many efforts are underway to define whether any given water project decreases or further increases water users' vulnerability to climate change, especially in the agricultural sector. A number of challenges in making these assessments have been routinely discussed, including a lack of yardsticks, varying local circumstances, and the role of subjective judgment. However, their common assumption is that maladaptation constitutes, at least in theory, a phenomenon that can ultimately be fully objectified and agreed upon through scientific data. By contrast, our qualitative comparative analysis advances a non-positivist conception of maladaptation. By comparing three water projects aiming at climate change adaptation in South Africa (Western Cape), Spain (Andalusia) and France (Occitanie), we argue that qualifying (mal)adaptation is an inescapably political process as much as a scientific endeavour. As a consequence, we advocate for putting more scholarly emphasis on the governance of adaptation projects, especially the capacity of governance arrangements to produce legitimate compromises between multiple policy domains and actors; to enhance higher level conflict management when necessary; and to ensure regular policy evaluation with an effective bearing on subsequent policy decisions.