The current design guidelines (e.g., ALA 2005) have been reported to underpredict the axial pullout resistance measured in laboratory and field tests for pipes buried in dense sand. The higher pullout resistances in the tests were believed to be due to the shearing-induced soil dilation at the pipe-soil interface. However, the mechanism of soil dilation could not be measured during the tests. In the current study, three-dimensional finite-element (FE) analysis was employed to examine the mechanism, which revealed that the effect of shearing-induced dilation could be insignificant, depending on the magnitude of the earth pressures. For pipes buried at shallow depths, the compaction-induced lateral earth pressures significantly contributed to higher interface normal stresses and the increase of normal stress due to shear-induced dilation, resulting in relatively higher pullout resistances. The stiffness of the pipe and soil also influenced the interface normal stress. The compaction-induced lateral earth pressure increase was modelled using equivalent temperature loads in the FE analysis that successfully simulated the measured pipe responses. Based on the findings, a modification to the current design equation to calculate the maximum axial spring force was proposed, incorporating the compaction-induced lateral earth pressure and a normal stress adjustment factor.