Growing rice is the main activity in northeast Thailand. Unpredictable droughts and coarse-textured soil are the principal constraints usually cited to explain the low yields and economic poverty of this region. Past studies tried to improve the drought tolerance of rice varieties and hydro-logical functioning at the field level. How water is used at the farm level remains largely unknown. Consequently, it is relevant to understand the interactions between the water-resource and water-use dynamics. This article proposes to develop a simulation tool based on multi-agent systems to explore adaptations of the rice cropping pattern to rain-fall variability. An environment containing the main biophysical entities involved in decision-making rules for water use is modeled and its hydrological functioning is verified. Preliminary simulations are run to illustrate the model capacities. The next step will consist of adding autonomous agents to simulate scenarios in which farmer agents may cooperate to use water.