The key areas of China's urbanization process have gradually shifted from urban areas to county-level units. Correspondingly, air pollution in county towns may be heavier than in urban areas, which has led to a lack of understanding of the pollution situation in such areas. In view of this, 236 PM2.5 filter samples were collected in Pingyao, north of the Fen-Wei Plain, one of the most polluted areas in China. Monte Carlo simulation was used to solve the serious uncertainties of traditional HRA, and the coupling technology of absolute principal component score-multiple linear regression (APCS-MLR) and health risk assessment (HRA) is used to quantitatively analyze the health risks of pollution sources. The results showed that PM2.5 concentration was highest in autumn, 3.73 times the 24 h guideline recommended by the World Health Organization (WHO). Children were more susceptible to heavy metals in the county-level unit, with high hazard quotient (HQ) values of Pb being the dominant factor leading to an increased non-carcinogenic risk. A significant carcinogenic risk was observed for all groups in autumn in Pingyao, with exposure to Ni in the outdoor environment being the main cause. Vehicle emissions and coal combustion were identified as two major sources of health threats. In short, China's countylevel population, about one-tenth of the world's population, faces far more health risks than expected.