Abstract: Based on the hourly data of carbonaceous aerosol, this study explored the characteristics, influencing factors, sources and potential contributing regions of carbonaceous aerosol concentration in Changji City in winter, using the random forest model, positive matrix factorization (PMF) and potential source contribution factor analysis (PSCF) methods. The results showed that the concentrations of organic carbon (OC) and elemental carbon (EC) on pollution days were 1.8 times and 1.4 times those on clean days, respectively, and their concentrations were lower in the day and higher at night. OC, primary organic carbon (POC) and secondary organic carbon (SOC) showed a significant bimodal distribution on pollution days, and SOC played a dominant role in the increment of carbonaceous aerosol on pollution days. The generation of SOC on pollution days was greatly influenced by the precursors and the environmental conditions. The PMF source analysis showed that the main sources of carbonaceous aerosols in Changji City were biomass/coal combustion (42.6%), gasoline vehicle emissions (26.8%), diesel vehicle emissions (18.4%) and road dust (12.2%). Biomass combustion and gasoline vehicle emissions contributed more prominently on pollution days. The PSCF results showed that the potential sources of carbon components on clean days were distributed in the local southwest direction, OC concentration on pollution days was mainly affected by the local and southwest direction, and EC and POC were affected by multiple long-distance transport sources.