The GHOST method is well known proximity effect correction scheme which is simple and accurate, and requires no extensive computer calculations. Although the tolerances on dose and beam diameter of correction exposure have been studied by several researchers, there has been no systematic evaluation of the tolerance on alignment between primary and correction exposure. In this research, the tolerance on alignment has been calculated for a two-dimensional test pattern by using a simple numerical calculation with the double-Gaussian approximation and a threshold energy model. The calculation shows that an alignment error causes much larger deviations in linewidth than in position shift. The tolerance on alignment, therefore, is determined by linewidth deviation. The tolerance has linear dependence on both the range of backscattered energy distributions (B(b)) and the ratio of the absorbed energy deposited by the backscattered electrons to the absorbed energy by the forward scattered electrons (eta). The larger B(b) and lower eta values provide greater tolerance to misalignment. The feasibility of applying the GHOST method to electron-beam energies of 10 keV, a range from 20 to 30 and 50 keV is discussed using the calculated results.