To ensure the soil pressure accuracy of trapdoor tests, according to different depths of the soil, a series of calibration experiments with different levels was conducted by using resistance strain-type soil pressure cells. The calibration included loading calibration and unloading calibration. Additionally, a modification was derived to the applied stress (i. e., effective stress) by considering the influence of boundary friction force. Through discrete element simulation (DEM), it is proved that the modification of applied loading stress is appropriate. The calibration test results reveal that the macrostrain-effective stress curve of loading progress exhibits linear variation and can be well-fitted by a linear function. The macrostrain-effective stress curve of unloading progress exhibits an evident hysteresis phenomenon. With increasing loading pressure, the hysteresis phenomenon becomes more significant. And it can be fitted by using a single exponential curve. After calibration, the filling validation were conducted. Results indicate that the soil pressure cell should not be used under low stress or shallow fill depths.