In this paper, three aspects (a total of four evidences) of controlling the fracture density of the hydrothermal transport channel, determining the source of ore- forming geochemical factors (anomalies, background), and reflecting the satellite gravity distribution of ore- forming surrounding rocks are significant for gold deposits. Starting with indicative prospecting indicators, the best prospecting evidence is obtained after testing with existing gold deposits, and for the first time, the weight of evidence model is used to calculate the posterior probability of gold metallogenesis across the country. The number of undiscovered gold deposits is estimated in the three probability intervals of 10%, 30% and 65%. Combined with the tonnage- grade model of gold deposits in North Africa, the national reserves and resources of Eritrea gold mines are estimated in three intervals. Based on the calculated posterior probability distribution, three- level favorable prospecting areas are divided according to the probability of finding gold deposits from high to low. The results show that Eritrea has good prospecting potential for gold deposits. Except for the traditional NNE and NNW trending gold metallogenic belts, the northern junction of the two and the southern interconnection area need to be paid attention to. The comparison with the existing metallogenic belts of Eritrea from traditional field geological surveys shows that the prospecting prospects of this division not only include the traditional gold metallogenic belts, but also found in the southern part that have not attracted attention before. The belt- shaped area provides a new direction for exploration in the Eritrea gold deposits. Various evidences also show that the junction of the two metallogenic belts also deserves more attention. Compared with Eritrea’s gold reserves and resources data obtained by other methods, this estimation result provides the results of Eritrea’s gold reserves and resources in different probability spaces for the first time, which provides more accurate data support to further understand the potential of Utt’s gold mine.