Abstract: Gas content is the most important research content for coalbed methane (CBM) development. Accuracy characterization is very important for CBM. Logging characterization of gas content is the most commonly used method, while different logs have different sensitivity to the change of gas content, and single logging curve predicting coalbed methane content has the disadvantages of poor stability and low accuracy. In order to study the accurate prediction method of CBM, CBM reservoirs in S Block of Australia is taken as study object. Based on the laboratory analysis data and logging data, the logging data response analysis was carried out to realize logging data expansion correction and depth shift. Logging data such as buried coal depth, AC, GR and LSD were selected as the basic parameters for gas content prediction on the basis of correlation analysis between gas content and logging data. After the sensitivity analysis between gas content and logging data, compound parameters of gas content prediction were constructed, and finally the compound parameter prediction model for gas content prediction was established. Batch calculation of CBM was achieved by plug-in module of logging interpretation software. In the practical application, the prediction model can overcome the shortcomings of the low accuracy and poor stability of the traditional coal seam gas content calculation. At the same time, it can realize the batch calculation and greatly speed up the calculation of gas content. The prediction of gas content by compound parameters can also provide a solid geological foundation for the CBM subsequent development of S block.