Аннотация:Remote assessment of soil salinity in natural solonetzic complexes, which are characterized by subsurface salinization of soils, is a difficult task. However, research in this area is promising, since salinity is a clear limiting factor that influences vegetation growth; thus, it affects the spectral characteristics of vegetation. This study carried out an analysis of multi-temporal high-resolution space images, which consisted of comparison with detailed ground-based data on soil salinity using the principal component method and multiple linear regression. Images from the QuickBird (2007) and SuperView-1 (2021) spacecraft with a spatial resolution of 2 m were used as remote sensing data. Ground studies were carried out in 2011 and 2021. Soil salinity was estimated from the specific electrical conductivity (EC) in an aqueous suspension of 1 : 5. It was found that there were no significant changes in soil salinity in the key area over the 10-year period; however, there were changes in the state of vegetation, which are reflected on the maps of the NDVI vegetation index. Multi-temporal high-resolution satellite images were used as a basis to calculate the principal components; it was concluded that the first three components explained almost 97% of the entire image variance. Models built using multiple linear regression analysis describe soil salinity well (R 2 of the model is 0.68, 0.77, 0.83 for the 0–30, 0–50, 0–100 cm layers, respectively). When tested on a control sample, the constructed models based on remote data showed good convergence (R 2 between the predicted and real values of EC is 0.70, 0.87, 0.83 for the 0–30, 0–50, 0–100 cm layers, respectively). The proposed models will be useful for assessing the salinization of soils in the solonetzic complex in the south of the steppe zone according to high-resolution satellite imagery.