Abstract: Aiming at the existing group multi-criteria sorting studies that do not consider the behavioral characteristics of experts and the complexity and time-consuming of reaching consensus on group reference profiles, this paper proposes a new group multi-criteria sorting method. Firstly, the hesitant fuzzy sets and the hesitant fuzzy linguistic term sets are used to portray the hesitant behavior of experts in giving reference profiles. Secondly, a mathematical planning-based method is used to reach consensus on group reference profiles based on the consideration of different adjustment amounts of different experts. Thirdly, the prospect theory is used to portray the bounded rational behavior of experts, and the alternatives’ classes are determined based on the comprehensive prospect value compared to group reference profiles. Finally, this method is applied to specific cases, and the effectiveness and superiority of this method are verified through comparative analysis. This method considers the behavioral characteristics of the experts in the sorting process, which is closer to the actual situation; in addition, using group reference profiles which are reached consensus to sort helps to obtain more reasonable and reliable sorting results.