The aim of this paper is to suggest the adoption of an "evidence based" approach in the analysis of CAP expenditure. The debate on CAP expenditure is generally based on ex-ante images of the budget as presented in the Multiannual Financial Framework and, before each financial year, on the budget appropriations for commitments. But this differs significantly from the actual payments as recorded ex-post in the Financial Reports. These differences are mainly concentrated on structural policies (such as regional and rural development policy), while they do not exist, or are minimal, in the mere transfer policies (such as in Pillar 1 of the CAP). Based as it is on a partial and distorted image of the expenditure, the analysis and discussion on the CAP reform is distorted as well.