The EU banana market has been of enormous interest for researchers for a long time, especially after the import policy unification brought by the Common Market Organization for Bananas (CMOB) in 1993. Empirical evidence suggests that the CMOB and its subsequent modifications have been highly distorting. However, the quantifications made of those distortions by different authors vary a lot, not only in magnitude but also in direction. The reason is that for each evaluation, a different system of demand equations has been estimated. However, besides the different models used, there is one common denominator to all models. They do not incorporate the general restrictions (homogeneity, symmetry and adding up) necessary to make the demand estimations consistent with economic theory. In this paper we estimate the almost ideal demand system (AIDS) to calculate more reliable elasticities to facilitate future welfare analysis of the EU banana market. The inverse almost ideal demand system (IAIDS) was also estimated under the alternative assumption that import prices adjust to quantities instead to quantities adjusting to price variations as it has usually assumed. However, the results are not reported because due to the upcoming import policy and its subsequent quota elimination, quantities will not longer be predetermined. Preliminary results show that demand elasticities are different depending on the region of origin of the imported bananas. Latin American bananas are substitutes of bananas from preferred suppliers and complements of others while imports from the ACP are substitutes of all regions. EU bananas behave similar to Latin American in the sense that they are substitutes of all by other suppliers. Preliminary welfare analysis of the proposed policy shows that both Latin American producers and EU consumers of Latin American bananas will lost from the quota elimination and increase in the tariff level.