We investigate the impact of mergers on the mass estimation of galaxy clusters using $N$-body + hydrodynamical simulation data. We estimate virial mass from these data and compare it with real mass. When the smaller subclusters mass is larger than a quarter of that of the larger one, virial mass can be larger than twice of the real mass. The results strongly depend on the observational directions, because of anisotropic velocity distribution of the member galaxies. We also make the X-ray surface brightness and spectroscopic-like temperature maps from the simulation data. The mass profile is estimated from these data on the assumption of hydrostatic equilibrium. In general, mass estimation with X-ray data gives us better results than virial mass estimation. The dependence upon observational directions is weaker than in case of virial mass estimation. When the system is observed along the collision axis, the projected mass tends to be underestimated. This fact should be noted especially when the virial and/or X-ray mass are compared with gravitational lensing results.