The paper shows the practical use possibilities of satellite services "Vega"for arable land use determining. The services provide access to satellite data and different tools for its processing for solving this problem. The article shows examples of agricultural land use types interpretation using different approaches such as multi-temporal RGB composites analysis and NDVI vegetation index dynamics analysis. The results of object-based classification of arable land use assessment are demonstrated in the paper on the example of Safonovsky District (Smolensk region, Russia). It is shown that the reliability of the proposed approach reaches more than 90%. © Published under licence by IOP Publishing Ltd.