Abstract: The soil arching effect is a redistribution of the stress in soil. It is caused by the rigidity difference between the soil and the structures in the soil. The studies so far about the impact of cyclic loading on the soil arching effect is very limited. A custom-made apparatus is used to conduct plain-strain trapdoor model tests under soil self-weight, static loading and cyclic loading to study the impact of different loading conditions on the soil arching effect. The analogical soil of aluminum rods is used as the fill instead of sand. The parameter of the soil arching ratio is used to evaluate the soil arching effect. The test results from this study are compared with those from the previous studies. It is indicated that both static loading and cyclic loading reduce the existing steady soil arching. The reduction of the soil arching effect increases with the loading magnitude and frequency and decreases with the loading area. Under the same loading level, the cyclic loading reduces the soil arching effect more than the static loading. The difference of reduction caused by them decreases with the loading magnitude and increases with the loading frequency. Besides, the difference under the peak pressure is less than that under the zero pressure. On the whole, the expressions given by Evans¹⁸ can well predict the soil arching ratios in the trapdoor tests under soil self-weight and static loading. However, an improvement of the expressions is necessary for the trapdoor tests under cyclic loading.