Abstract: In order to make scientific decisions of suppliers under the global value chain environment, a multi-criteria group decision-making model based on uncertain linguistic terms was proposed. Preference information, first elicited from experts, was transformed into hesitant fuzzy linguistic terms and computed with words via uncertain linguistic variables. The expert group's preference information was then fused by the envelope operator and hesitant fuzzy linguistic term set built. Then the relative closeness coefficient was adopted to sort the production suppliers. Consequently, the most satisfactory supplier was selected. In addition, information entropy was proposed for solving the weights without prior knowledge of multi-criteria decision-making process. The results show that the most satisfactory selection results are the same under 3 different information entropy parameters. Moreover, the ranking results of the relative closeness coefficient for supplier selections are not sensitive to the change of information entropy parameters, which verifies the feasibility, effectiveness and stability of the proposed model. Therefore, the proposed model can provide a useful reference for the practical application of auto parts suppliers' evaluation and selection.