Satellite images Landsat are suitable for spatial and temporal analyses since they have a wide area coverage and a huge archive that is freely available. In this thesis, we present changes in land cover in an area located in the south-western Slovenia throughout a long-time period. We used six Landsat satellite images, taken between 1986 and 2016. For these images, we performed supervised classification. With the method of spectral angle mapper algorithm, we classified land cover into five classes (built-up area, forest, water, fields and meadows, and open areas). We evaluated the results of individual classification and discussion of individual results. The analysis of the change in land cover is based on the results of image classification, made with a difference of twenty-five years. The changes in built-up areas, forests, fields and meadows, and open areas are shown and discussed.