Coltan is a strategic mineral as it is one of the most important ores for the extraction of tantalum and niobium, two resources with very special physicochemical properties and of enormous interest for technological development from the 20th century (Hughes et al. 2011). The concentrations of uranium and thorium in Venezuelan coltan samples, previously measured by the Simón Bolívar University Nuclear Physics Laboratory (LFN-USB) and the previous description of several alluvial deposits in the Parguaza Region (southern Venezuela) showed the potential feasibility of localizing shallow coltan deposits using gamma-ray spectrometry by mapping the radiation associated with U and Th in these minerals. In this work we investigate this probability by using a high-density scintillation detector (NaI 4”x4” Bicron 4H4/3). Coltan samples were used on a test field conditioned to emulate a shallow alluvial deposit, as those observed in some areas of the Parguaza region, obtaining that for detailed prospecting a 2.5x2.5 m2 spaced survey grid using a BGO detector is effective in detecting significant quantities of coltan (kilograms), even at a depth of 40 cm (depth at which artisanal extractions have been described in recent years). The need to prioritize the quantification of thorium and especially uranium was also evidenced, using their more energetic emission lines. This first approach for the prospection of coltan showed promising results with data processing allowing a clear identification of coltan location. The high energy gamma emissions of U (1765 keV) and Th (2615 keV) have a high penetrating power so they can easily pass through several centimeters (near to a meter in the tested cases). This is why the proposed method is rather adequate in a context with many scattered spots of some kilograms of coltan each, buried up to 1 m underground. Since most of the low energy emission of the gamma spectrum (also the more frequent from zones without coltan) will be attenuated (filtered) by the ground, the quantification of the total-count has been confirmed to be a good classifying method for larger formations or different geological areas. While on the other hand, a detailed high special resolution survey focused on the detection and quantification of U and Th, is a way to track down coltan sources, resulting in the location of what artisanal miners describe as “sweet spots”.