In this paper, the authors take Cu element of Tongling ore clustering area as an example, and carry out the research on metallogenic anomaly identification with multifractal method, based on deep soil data. The results show that the Krige-multifractal interpolation method has tremendous advantage compared with the Krige interpolation method when soil sampling density is relatively lower and spatial distribution of element content has little difference. For the kind of old ore clustering area, such as Tongling ore clustering area, which has a higher level of ore mining, and the spatial distribution of surface soil elements is mainly controlled by external pollution sources, the multifractal method that is based on deep soil data can identify anomalies efficiently and provide special services for deep and peripheral concealed ore prospecting in old ore clustering area.