Chloride ions will corrode the concrete structure, cause the corrosion of the reinforcement inside the structure, lead to concrete cracking, and destroy the structures integrity. Conventional chlorine content measurement methods, such as chemical titration, have many problems, such as complicated operation and slow detection speed. Laser-induced breakdown spectroscopy (LIBS) has the advantages of no sample pretreatment required, two-dimensional scanning detection and in-situ fixed-point analysis. However, there are still some problems in the measurement of chlorine content in concrete, such as the difficulty in reducing the limit of chlorine quantification and the influence of non-cement components on the measurement results of the two-dimensional distribution of chlorine. In order to meet the application requirements of rapid detection of chlorine content in concrete structures of nuclear power plants, a method for detecting chlorine content distribution by dual-pulse LIBS is studied,and the erosion status of chloride ions in simulated concrete samples of nuclear power plants is evaluated in the paper. Firstly, the calibration models of chlorine are established by internal standard (IS), Principal Component Regression (PCR) and Support Vector Regression (SVR). The LOD and LOQ of chlorine calculated by the IS method are 0.006 02 wt% and 0.0180 6 wt% respectively. The Leave-One-Out Cross-Validation (LOOCV)method is used to evaluate the prediction performance of the three calibration models. Secondly, to exclude the influence of the non-cement matrix on chlorine detection, Logistic Regression combined with Principal Component Analysis (PCA) and SVM classification are established to identify aggregate and cement in concrete. The SVM model with the combination of Si, Ca and O has the best classification effect, and its recognition accuracy of the total components reaches 98.20%, including 97.84% for aggregate and 98.33% for cement. Finally, the quantitative analysis of chlorine in the concrete erosion surface corroded for 15 and 30 days iscarried out aPCR calibration model. The average predicted chlorine content of the two concrete samples reach the maximum value at about 2 mm, 0.890 wt% and 0.599 wt%, respectively, which is consistent with the result of the potentiometric titration method. In conclusion, based on the dual-pulse LIBS method, chlorines LOD the LOD of chlorine comes to 0.006 02 wt% under the total energy of 30 mJ. The precise identification of aggregate and cement in concrete is realized, and the two-dimensional distribution of chlorine content on the erosion and penetration surface of concrete is obtained, which provides an engineering solution for the field application of rapid detection of chlorine content in nuclear power plant concrete structures.