This paper presents a methodological approach to analysing combinations of different agricultural policy and irrigation water pricing alternatives. The approach takes into account the fact that farmers consider a broad set of criteria when they make decisions. Policy scenarios are simulated by means of multi-criteria mathematical programming models capable of simulating the future behaviour of farmers. The Multi-Attribute Utility Theory (MAUT) has been chosen as a methodological framework for model-building. A set of relevant economic, social and environmental attributes related to public criteria are also obtained as a way of measuring the efficiency of the suggested policy scenarios. This methodology has been applied to the Duero Basin in Spain.