A method of axially aligning a charged particle beam involves an image data acquisition step and a calculation step. The image data acquisition step consists of obtaining first to third sets of image data by scanning a shielding member placed in the path of the beam with the beam while varying conditions of the excitation currents through first and second alignment coils, respectively. The calculation step consists of calculating the values of the excitation currents through the first and second alignment coils, respectively, for axial alignment of the beam, based on the obtained first to third sets of image data.