Purpose. The research purpose is to study the forms of gold occurrence in the Takyr graben alluvial-proluvial deposits in order to develop a low-waste resource-saving processing technology. Methods. At the first stage, in the course of field studies, the geological structure of the site is specified, samples are taken for mineralogical-technological analysis, which includes the following procedures: studying granulometric and mineral composition of detrital material; fractional gravitational beneficiation of the source material; fractional beneficiation of material pre-processed in the autogenous mill (AG mill); studying free (native) and bound gold in beneficiation products, their quantitative assessment. The samples are processed under the three-stage scheme using an autogenous mill for sample preparation. Findings. For the first time, the geological structure of the site has been specified with the identification of the deposits in the Turangi and Tuzkabak suites within its boundaries, as well as granulometric and petrographic composition of gold-bearing deposits, and the specifics of gold bearing. The detrital material, represented by quartz with a sharply subordinate amount of quartzite, quartz diorites and jasperoids, is practically identical in all fractions. Originality. For the first time, various types of gold mineralization have been identified in placer sands: clastogenic, newly-formed hypergene, newly-formed hydrothermal and residual. Fractional beneficiation makes it possible to estimate the gold content in each fraction and gravitational beneficiation products, as well as the ratio of free and bound native gold in different fractions. The largest amount of free native gold has been revealed in fractions of -0.25 + 0.1 mm (60%) and -0.074 + 0.044 mm (~40%). Gold is high-grade (96.5%) with an admixture of silver and iron. Together with gold, ilmenite, zircon, scheelite, native bismuth, as well as barite, galena, sphalerite, and dolomite have been identified. Practical implications. The research results make it possible to reassess the prospects of similar objects, to adjust the scheme and methodology for processing stream-sediment samples, to solve the issues of productive sand processing technology, as well as to improve the efficiency of geological exploration and eliminate the “underestimation” of gold deposits. The results obtained can be recommended for implementation by both domestic and foreign organizations specializing in the exploration and mining of gold deposits.