Given the advantages of remote sensing, an increasing number of satellite aerosol optical depths (AOD) have been utilized to evaluate near-ground PM_2.5. However, the spatiotemporal relationship between AODs and PM_2.5 still lacks a comprehensive investigation, especially in some regions with severe pollution within China. Here, we investigated the spatiotemporal relationships between several satellite AODs and the near-surface PM_2.5 concentration across China and its 14 representative regions during 2016–2018 using the correlation coefficient (R), the PM_2.5/AOD ratio (η), the geo-detector (q), and the different aerosol-dominated regimes. The results showed that the MODIS AOD from the Multi-Angle Implementation of Atmospheric Correction (MAIAC) algorithm strongly correlates with PM_2.5 (R > 0.6) in China, particularly in the Chengyu (CY), Beijing-Tianjin-Hebei (BTH), and Yangtze River Delta (YRD) regions. The close correlations (R = 0.7) exist between PM_2.5 and MODIS and VIIRS AOD from the deep blue (DB) algorithm in the CY, BTH, and YRD regions. Under the key aerosols affecting China (e.g., sulfate and dust), there is a strong correlation (R > 0.5) between the PM_2.5 and MODIS and VIIRS AODs from the MAIAC and DB algorithms, with the higher concentration of ground-level PM_2.5 per unit of these AODs (η > 130). The MAIAC AOD (Terra/Aqua) can better explain the spatial distribution (q > 0.4) of PM_2.5 than those of AODs from the dark target (DT) and DB algorithms applied to the MODIS over China and its specific regions across seasons. The performance of the Advanced Himawari Imager (AHI) AOD (R > 0.5, q > 0.3) was close to that of the MAIAC AOD during the spring and summer; however, it was far less than the MAIAC AOD in the autumn and winter seasons. The investigation provides instructions for estimating the near-surface PM_2.5 concentration based on AOD in different regions of China.