The timely and effective evaluation of the effect of urban expansion on PM2.5 concentrations is highly important for an understanding of urban-environment sustainability in China. However, studies examining the effect of urban expansion on PM2.5 concentrations at multiple administrative scales from different perspectives are still lacking. In this study, multisource remote sensing data were used to define urban spatial expansion and urban socioeconomic expansion, respectively. The effect of urban expansion on PM2.5 concentrations was then evaluated and compared at the regional, urban agglomeration, and metropolitan scales. The results showed that China's economic urban expansion lagged far behind urban spatial expansion from 2000 to 2015. There is a bidirectional long-term Granger causal relationship between urban expansion and PM2.5 concentrations across different scales. A short-term Granger causality test proved that urban expansion is still at the stage of aggravating PM2.5 concentrations. Variance decomposition analysis results indicated that the contribution of urban socioeconomic expansion in explaining PM2.5 concentrations was higher than urban spatial expansion with the increase in scales. The aggregation effect has become an important means for urban socioeconomic expansion to affect PM2.5 concentrations. In addition, PM2.5 concentrations can be reduced by the technical progress effect of urban spatial expansion across different scales. The Chinese government should consider the impact of different urban expansion modes on PM2.5 concentrations. Moreover, the technical progress effect will play an important role in decoupling urban spatial expansion and PM2.5 concentrations in China.