The analysis has been conducted of possibility of determining the extent of soil salinity by indirect and direct decoding of satellite images Pleiades 1A / 1B (in the year of survey) and LANDSATTM (archive). It was found that indirect decoding of salinity based on spectral char-acteristics of vegetation images is strongly dependent on crop growth phase (period of shoot-ing). LANDSAT archive images with normalized soil salinity index (NDSI) allow to develop salinity maps and soil salinity dynamics maps at semi-quantitative level. Based on computer analysis of LANDSAT images it was determined that soil salinity at study object during the pe-riod from 1987 to 2014 has increased due to significant decrease of the area of non-saline soil by 41.5% and increasing of the areas of low and moderately saline soils by 34.9%, also regions with heavily saline soils at 6.6% of total surveyed area have been detected, which previously were absent.