We present our study on a comparison of dark matter halo merger history from the runs using different numerical simulation codes. To analyze the uncertainty caused by the use of different N-body calculation methods, we compare the results from two cosmological hydrodynamic simulation codes GADGET-2 and RAMSES, which use a TreePM algorithm and the Adaptive Mesh Refinement(AMR) technique respectively. We perform cosmological dark matter-only simulations with the same parameter set and initial condition for both. The dark matter halo mass functions from two simulation runs correspond well with each other, except for lower mass haloes. The discrepancy on the low-mass haloes in turn causes a notable difference in halo merger rate, especially for the case of extremely minor merger. The result from GADGET-2 predicts that most haloes undergo more number of mergers with small haloes than that from RAMSES, independent of halo mass and environment. However, in the context of the study on galaxy evolution, such extreme minor mergers generally do not have strong effects on galaxy properties such as morphology or star formation history. Hence, we suggest that this uncertainty could be quantitatively negligible, and the results from two simulations are reliable even with only minor difference in merger history.