In order to evaluate shale gas content more accurately, a novel method for estimation of shale gas content in place is proposed based on the Polanyi adsorption potential and the London dispersion potential energy (P-L method). A series of integrated analyses were conducted on one shale core sample from the 7th member of Yanchang Formation (Chang 7 Member) in Ordos Basin, NW China, including on-site canister desorption tests, high-pressure methane adsorption isotherms, helium-based porosimetry, and gas composition analysis. On the basis of our experimental results, the P-L method and the commonly used 'direct' (based on measurement of desorbed gas) and 'indirect' (based on measurement of pore volume and state of equation) methods were used, respectively, to model the in-situ shale gas content. The results show that the shale gas content under the same experimental conditions predicted by the P-L method was 4.398 cm(3)/g and that by the direct method was 1.668 cm(3)/g, which is significantly lower. Meanwhile, those under the same lab conditions by the P-L method and the indirect method are 3.954 cm(3)/g and 3.820 cm(3)/g, respectively. Direct methods are more sensitive to human errors than the indirect methods, remove the impact of arbitrary factors by operators, and can shed light on the actual shale gas content in place based on the measured properties of samples. The P-L method is supported by well-developed theoretical basis. Through comparison between all these three methods, the results by the P-L method can be reasonable and practicable, and could be used to recover shale gas content in-situ during the shale gas resource potential assessment.