Abstract: Coal bed methane content is a key technical parameter for mine gas prevention and control. In the process of direct determination of coal bed methane content sampling, due to the existence of wind pressure in the borehole, the gas desorption of coal samples in the borehole is in a pressure environment. However, at present, the gas loss of coal samples in boreholes is calculated based on the gas desorption amount under normal pressure environment, which results in inaccurate determination of coal bed methane content. Based on this, this paper developed a set of coal gas desorption experimental platform under pressure environment, explored the evolution law of gas desorption amount, desorption rate and diffusion coefficient of coal samples under positive pressure and variable pressure environment, constructed a coal gas desorption model under pressure environment, and used finite element analysis software to numerically solve the diffusion coefficient equation. The results show that the gas desorption process of coal samples is in a positive pressure environment during the migration stage in the hole, and the desorption amount, desorption rate and diffusion coefficient are negatively correlated with the desorption pressure. In the orifice sampling stage, the coal sample changes from the pressure environment to the normal pressure environment, and the gas desorption amount, desorption rate, diffusion coefficient and gas pressure show a positive correlation in the normal pressure environment. The loss calculated by the gas desorption model of coal under pressure is closer to the gas desorption amount of the experimental test. However, the gas loss calculated by the traditional method (\sqrt t method and power function method) is far from the gas desorption amount of the experimental test. The traditional method needs to measure the desorption amount to indirectly calculate the loss, and the pressure desorption model can directly obtain the loss according to the sampling wind pressure, gas pressure and sampling time. The research results lay a theoretical foundation for the accurate calculation of gas loss and the prevention and control of coal seam gas disaster.