Watershed planning and management is a complex process due to existing different influential criteria in social, economic and environmental sectors as well as complicated interactions among them. Therefore, evaluating watershed management programs needs a comprehensive approach considering network relations among criteria and diverse decision-makers’ judgments. This paper presents a new framework of group decision-making in evaluating and ranking watersheds for implementing development strategies. The model includes 18 criteria in social, economic, managerial and environmental clusters. Five watersheds are considered as alternatives. At first, the network relations among the various criteria are determined using the Interpretive Structural Modeling (ISM) method. Then, based on the relationships obtained by ISM and decision-makers’ judgments, the alternatives’ priority is determined through the Fuzzy Analytic Network Process (FANP) method. To aggregate the group decision-making results, the extended fuzzy TOPSIS method is proposed and the decision-makers’ weights are calculated using Shannon’s entropy method. Finally, the sensitivity analysis of the decision-makers’ weights and alternatives priorities have been conducted. The Monte Carlo method is applied to generate data for sensitivity analysis. According to the results, Urmia Lake is the most preferred watershed and Atrak watershed is ranked in last.