For accurate counting of alpha tracks on the polyallyl diglycol carbonate of CR-39-type track detectors,the size distributions of both artifact tracks and alpha tracks were investigated with an automatic counting system. At the same temperature and etchant concentration, the numbers and sizes of alpha tracks changed significantly with the etching time, and the artifact track changes were smaller. At the etching time of 5 h, the sizes of alpha tracks were evidently larger than those of the artifact tracks, and the deviation of its size distribution was much smaller than those of longer etching time. Based on the size distribution of alpha tracks etched for 5 h, the overlap effect and uncertainty of overlap correction were studied by the Monte Carlo simulations for different track densities. It was found that the counting uncertainty of the system could be less than 6% in a density range of 10–160 tracks mm~(-2) after taking the overlap correction into account. For accurate counting of alpha tracks on the polyallyl diglycol carbonate of CR-39-type track detectors, the size distributions of both artifact tracks and alpha tracks were investigated with an automatic counting system. At the same temperature and concentration, the numbers and sizes of alpha tracks changed significantly with the etching time, and the artifact track changes were smaller. At the etching time of 5 h, the sizes of alpha tracks were evidently larger than those of the artifact tracks, and the deviation of its size distribution was much Based on the size distribution of alpha tracks etched for 5 h, the overlap effect and uncertainty of overlap correction were studied by the Monte Carlo simulations for different track densities. It was found that the counting uncertainty of the system could be less than 6% in a density range of 10-160 tracks mm ~ (-2) after taking the overlap correction into account.