The study on the efficiency of transport-driven poverty alleviation can provide an important theoretical basis for the formulation of regional poverty alleviation-oriented transport policies in the stage of rural revitalization. Panel data of transportation and economic development in 56 counties in the border mountainous areas of western Yunnan from 2016-2019 are used as samples to construct a transport-driven poverty alleviation efficiency input-output index system. The DEA model and Malmquist index are used to measure the transport-driven poverty alleviation efficiency of counties， and analyze their spatial and temporal distribution characteristics. At the same time， Tobit regression model is constructed to explore the factors influencing the efficiency of county-level transport-driven poverty alleviation. The results show that： 1） The overall efficiency of transport-driven poverty alleviation in the mountainous counties along the border of western Yunnan shows a gradual growth. There is still large room for improvement in the stage of rural revitalization， and the comprehensive efficiency of transport-driven poverty alleviation is mainly influenced by the scale efficiency. 2） According to two dimensions of comprehensive efficiency average of transport-driven poverty alleviation and Malmquist index value， the 56 counties in western Yunnan can be divided into four types： potential， developmental， mature and consolidation types. The developmental and mature types are main types of efficiency for transport-driven poverty alleviation in the border mountainous areas of western Yunnan. 3） County size and industrial structure have a reverse inhibitory effect on the improvement of transport-driven poverty alleviation efficiency， while employment support， consumption level， location advantage， and investment scale all have a certain positive promoting effect on transport-driven poverty alleviation efficiency in the border mountainous areas of western Yunnan to different degrees. Each influencing factor shows differences in terms of its influence intensity and effect path on the county transport-driven poverty alleviation efficiency in different development stages. Therefore， we should further optimize the spatial layout of the county transportation infrastructure， attach importance to the development of digital and intelligent transportation， strengthen the “linkage effect” between labor employment and transportation development， and adopt differentiated and rational development strategies based on different development stages of transport-driven poverty alleviation and characteristics of influencing factors.