Data envelopment analysis is a useful model for efficiency evaluation. In order to improve the haze pollution in China, it is particularly essential to allocate the PM_2.5 emission rights of various provinces according to the actual situation reasonably. However, the expressions of input and output information are often limited or incomplete. Probabilistic linguistic term sets are employed to increase the expression range of information. Therefore, the proposed probabilistic linguistic term envelopment analysis (PLTEA) model is divided into deviation-oriented probabilistic linguistic term envelopment analysis model and score-oriented probabilistic linguistic term envelopment analysis model. The dual form of the PLTEA model has obtained to solve the actual emission rights distribution problem. Finally, the optimization methods of the distribution efficiency of the PM_2.5 emission rights in China in 2019, and some policy suggestions are given to reduce haze pollution.