A new pattern width evaluation method for electron-beam (e-beam) writing has been developed. The method can predict the overall pattern width accuracy objectively and quantitatively by taking into account the exposure intensity distribution function and specific e-beam writing data in which the dose is often modulated for proximity effect correction. The evaluation is carried out by calculating the effective exposure energies and distances for sampling points on layout patterns from the ideal location of the pattern edge, the probability of pattern edge resolution as a function of the distance, and the mean and standard deviation of the edge positioning accuracy. Experiments verify the validity of this approach. The method enables the evaluation of linewidth accuracy using e-beam writing data after proximity effect correction. Furthermore, it can be applied under various writing conditions and has the potential for use in the development of e-beam writing systems.