The Taklamakan Desert is one of the largest sources of mineral dust aerosols in Eurasia and exerts a strong influence on air quality and public health in surrounding regions. This study presents a comprehensive spatiotemporal assessment of PM2.5 and PM10 pollution across five prefecture-level administrative units surrounding the Taklamakan Desert (Korla, Aksu, Kashgar, Hotan, and Kizilsu) during 2015- 2024, using long-term observations from China's national air quality monitoring network. Seasonal variability, long-term trends, extreme pollution events, and population exposure were systematically analyzed. The results reveal persistently elevated particulate matter concentrations, with PM2.5 and PM10 levels exceeding World Health Organization air quality guidelines by approximately 15-20 times on most observation days, indicating chronic rather than episodic exposure. A pronounced and recurrent spring maximum dominates the seasonal cycle across all regions. Consistently low PM2.5/PM10 ratios ( similar to 0.35) demonstrate the predominance of coarse mineral dust and confirm the dominant role of natural desert emissions rather than local anthropogenic sources. Extreme pollution events occur frequently and often synchronously across multiple prefectures, reflecting large-scale dust outbreaks amplified by the orographically enclosed morphology of the Tarim Basin. Population-weighted exposure analysis reveals strong spatial inequality in health burden. Kashgar and Aksu experience the highest cumulative exposure due to large population size, while Hotan represents the most severe per-capita exposure hotspot owing to persistently extreme concentrations. Overall, the findings demonstrate that air pollution and associated health risks in desert-margin regions are primarily governed by natural dust processes and regional atmospheric conditions, underscoring the need for improved early-warning systems and adaptive public health strategies in arid environments. (c) 2026 China University of Geosciences (Beijing) and Peking University. Published by Elsevier B.V. on behalf of China University of Geosciences (Beijing). This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).