Abstract: Objective Mineral identification is the basis of geological work, but the traditional identification is greatly affected by the experience of personnel and has strong subjectivity. Method This article takes different types of rock light thin slice samples and single mineral samples as research objects, and uses laser Raman spectral mapping technology to conduct large-area spectrum scanning to study the practical application effect of this method in mineral identification. From the selection of experimental mode, the setting of experimental conditions to the final imaging, several factors restricting the practical application of this technology are analyzed around the entire testing process. The principle of setting experimental conditions is put forward, the application field of the technique in mineral identification is discussed, and some suggestions for future development are put forward. Result Compared with the traditional identification, the technology has a stronger ability to distinguish fine minerals and a higher accuracy. The chemical imaging map of its components can intuitively understand the composition and distribution of various minerals, and can also carry out statistics of various contents in single minerals. It is considered that sample type is the main factor affecting imaging, and experimental conditions are secondary factors, which should be determined before large-scale scanning. Among them, the delineation of the imaging area, the length of exposure time, the collection step interval, and the setting of the collection point should be considered in combination with the characteristics of the sample type. The test accuracy directly affects the imaging effect, but it is inversely proportional to the test time. In order to balance the needs of the two aspects, it is also necessary to combine the actual situation of the sample. In the face of Raman spectra containing multi-mineral mixed peaks, altered minerals, homomorphic and homologous minerals, the ability of spectral resolution and identification of personnel is also indispensable. Conclusion The analysis shows that the point model of the technique is suitable for single minerals with scattered particles. The linear model is suitable for layered samples, such as SLATE. The range model is more suitable for mineral samples with uniform composition, uniform particle size and small alteration. The technique, combined with the microscopic identification method, can provide a more accurate, intuitive and convenient research method for mineral identification.