This paper described the equipment and test methods of the two dimensional soil arching laboratory tests. The distribution of vertical stress in the embankment was reported, and compared with the results of Terzaghi and Hewlett & Randolph methods. The modification of the Hewlett & Randolph method was also discussed. It was found that the ratio of embankment height h and pile spacing s play an important role in the formation of soil arching. The laboratory test results show that there is no soil arching for h/s <0.67, partial soil arching generated for 0.67< h/s <1.33 and fully soil arching generated for h/s ≥1.33. The distribution of vertical stress in the embankment captures the same tend of the results of Hewlett & Randolph method for h/s ≥1.33. The difference between the results of modified Hewlett & Randolph method and experimental results is smaller than that of Hewlett & Randolph method. The effect of embankment height on the stress ratio between piles and subsoil n were also investigated, it was found that the value of n increases with increasing the embankment height, whereas the growth rate reduced with the increase of embankment height.