In 2017, the Central American Agricultural Council (CAC), with the support of CCAFS, CIAT, IICA, FAO, ECLAC and CATIE, formulated and adopted the Climate Smart Agriculture Strategy (EASAC) for the Central American Integration System (SICA) region. This regional strategy aims at promoting climate-smart agriculture (CSA) at scale and is an original example of regional policy. This communication presents the first results of an assessment process of the EASAC outcomes three years after its launch. We carried out a methodology mobilizing the theory-based assessment (Delahais and Toulemonde 2012; Lemire, et al. 2012, Maine, 2008) applied to a policy innovation, in this case, the EASAC. Theory-based approaches have been designed to provide systematic, robust approaches to understand whether the intended outcomes of an intervention have been achieved (or not), and the importance of the intervention's contribution under consideration, relative to that of other alternative causes. This assessment process encompasses three steps: (i) the formulation of the EASAC Theory of Change (ToC), (ii) the identification of changes aligned with the ToC and (iii) the analysis of EASAC's contribution to some of these identified changes. In this communication, we present the results of the two first above mentioned steps. To establish the EASAC ToC, we first reformulated the EASAC into a ToC vocabulary/thinking starting from the policy document (CAC, 2017) and adjusting it through an iterative process of consultation and validation with 8 key actors (EASAC formulators). Once the ToC validated, we conducted 44 semi-directed interviews with actors involved in CSA related actions or interventions at regional and national level of the 8 countries of SICA. The objective of these interviews was to understand what expected changes identified in the EASAC ToC have occurred so far. Finally, the changes were analyzed and compared across and among countries. The EASAC ToC is structured along four rout