We present a state-of-the-art of spectral interpretation of induced polarization (IP) data; we discuss the advantages and drawbacks of this approach, as well as perspectives of its further development. We briefly show how spectral parameters can be obtained from time domain IP data and how they can be applied in ore prospecting. Based on experience obtained in VIRG-Rudgeofizika Ltd., we show that the standard chargeability-based approach alone can lead to missing of perspective ore bodies or to overestimation of bodies without economic interest. In contrast, consideration of spectral data allows discriminate between these classes of potential drilling targets. We present two case studies of combined application of the chargeability-based approach and of spectral approach (two gold deposits: Maletovaiam, Kamchatka, and Yasnoe, Taimyr), which allows in combination to discover areas of prospecting interest (e.g., hydrothermally altered volcano-sedimentary rocks and berezitized rocks containing pyrite). We also highlight that a main complication in use of the spectral approach is not a time-consuming field measurements, but rather the time-consuming visual data analysis. We propose, therefore, a new algorithm based on the pattern recognition principle, and we applied it to the field data.