In recent years, four Nordic countries have implemented large-scale police reforms. Reforms have been evaluated and the present article discusses the merits of such processes. Three topics are discussed: first, it is argued that such evaluations are by nature of limited methodological rigour. The common model of nationwide implementation at one point in time limits possibilities for more sophisticated models of evaluation. Second, due to the nature of such evaluations, they are of limited value as a tool for organizational reorganization. In fact, organizational response to evaluation may take the form of denial rather than change. Third, it might be assumed that countries planning major police reforms would study such evaluations in detail and revise their plans accordingly, but available evidence from Norway and Sweden suggests otherwise. The article concludes with a discussion whether it would be possible to devise evaluations in ways that would increase their worth and impact.