Based on the meteorological data of Global Data Assimilation System (GDAS) and PM2.5 concentration data from the China National Environmental Monitoring Station of the Ministry of Ecology and Environment, using MeteoInfo and TrajStat plug-in, a cluster analysis on the airflow trajectory reaching the main urban area of Yining City, China, from December 2018 to January 2019 was carried out, and then the weighted potential source contribution probability (WPSCF) and the weighted concentration weight trajectory (WCWT) of PM2.5 in the main urban area of Yining City was analyzed by using potential source contribution function (PSCF) analysis and concentration weighted trajectory model (CWT) analysis. The results show that: a) Trajectory clustering shows that the region is most affected by the southwest air mass, mainly from Kazakhstan and Kyrgyzstan. b) The WPSCF and WCWT distribution feature are similar, and the highest WPSCF and WCWT in the heavily polluted source area are above 0.7～0.8 and 80～110 μg·m-3 respectively, which are mainly concentrated in Zhaosu-Tekes Basin.