The agriculture system is an economic and environmental compound system; consequently, the three major economic, social and environmental factors should be considered in the optimal water resources allocation along with corresponding crop pattern under the constraints of availability of land and water resources. An interactive fuzzy approach is applied to develop a sustainable crop pattern for solving multi-objective programming involving vague information related to data, model formulation, and the decision maker's preferences. Furthermore, a model of relative order degree entropy is formulated to evaluate the reasonability of crop pattern alternatives. The case shows that the optimization and evaluation models for multi-objective crop pattern presented in this paper are suitable, and can supply theoretical references for a sustainable crop pattern and irrigation water resources allocation.