The aim of this work was to evaluate the capabilities and feasibility of using the averaged mass spectra obtained from pyrograms for assessing the group distributions and similarities of vehicle paint samples.To achieve this, all the profiles of averaged mass spectra were examined using hierarchical cluster analysis and principal component analysis after data processing. The pyrograms of 54 vehicle top coating samples were characterized by mass fragments chosen from 15 major paint constituents commonly used in the paint industry and multivariate chemometric analysis based on 76 variables, used as indicators for the classification of paint samples. Not only the characteristics of the variables were discussed, but the discriminating power among the samples was also evaluated by adding six blind test samples. In addition, validation of this method was accomplished by mixing three sets of international proficiency test samples and 23 randomly selected paired samples for comparison purposes, and the results of these gave the expected discriminating capability of the technique developed. The results indicated that this method was a promising, efficient and time saving tool for examining top-coating paint traces, with the application of combined pyrolysis-gas chromatography/mass spectroscopy with multivariate chemometrics for forensic purposes.