The article dwells upon the effective method of seismic data analysis called spectral decomposition. This method has been used worldwide for identifying additional geological features, as a tool for seismic geomorphology in the conditions of complicated geological structures during identification of paleochannels, lenses, and thin formations. It should be noted that spectral decomposition is not a direct tool for identifying the hydrocarbon saturation. It allows to represent the seismic data in the RGB format (red, green, blue) by way of decomposing seismic data into frequency elements. Authors demonstrate results of applying the seismic data spectral decomposition method within the terrigenous deposits on the example of gas fields in the Carbon age within the northern edge of Dnieper-Donets Depression. Based on the results of the available seismic material analysis, it has been shown that within thick and porous sand bodies, spectral decomposition demonstrates reasonable results, and the most informative frequencies are 25, 28, 31, 37, 39, 47, 56 Hz. By confirming the really existing deposits in the Moskovian and Bashkirian ages, spectral decomposition gives grounds for considering this method an effective tool to be used within the northern edge of the DDD, because the latter usually clearly responds to lithology, and sometimes even to saturation. The 8 most representative horizons have been analyzed: М-2а, М-3а, М-4, М-6, М-7, B-6 (at different stratigraphic levels), B-7 and S-6. The most definite results have been received within the following horizons: M-2a (the deposit has been confirmed by drilling), M-3a (the deposit and paleochannel have been confirmed by drilling), M-4 (paleochannel). Having regarded the spectral decomposition in complex with the results of other geophysical methods, some perspective objects have been identified. The bodies have been confirmed by the results of AVO and Vp/Vs analyses. Both objects served the targets for two prospecting wells located in the most optimal conditions. It has been brightly illustrated that the spectral decompositions as a tool for geological prospecting should be specifically used when some individual lenses and channel bodies are being prospected, with structural factor being present