Integrated floating photovoltaic-pumped storage power (FPV-PSP) system provides a promising way to solve the instability of photovoltaic output and the shortage of land resources. Site selection is the primary issue to the success of the FPV-PSP system, whose complexity generally comes from the uncertainties of decision-making environment, differences in large-scale group evaluation information and limited behavior characteristics of decision-makers (DMs). To this end, a large-scale group decision-making framework is constructed in this paper. Firstly, nineteen influencing factors of FPV-PSP system site selection are identified and classified into four groups. The probabilistic linguistic term sets are then introduced to evaluate the criteria values to depict the uncertainties. Afterward, a new probabilistic linguistic similarity measure is defined to divide DMs into groups, so as to aggregate the different large-scale group opinions. Thereafter, to simulate the psychological behavior of DMs, an extended fuzzy PROMETHEE method (Preference Ranking Organization Method for Enrichment Evaluation) is proposed to rank alternatives. Finally, to validate the effectiveness of the proposed framework, a case study in China is conducted and the calculation results show that the system situated in Lianyungang (A2) is the optimal location and the ranking order varies with the changes of regret aversion coefficient.