High-functional plastics have been attracted much attention over the various fields of industries. For evaluating their functionalities, it is essential to investigate low molecular weight components on the surface of the plastic because these components deeply affect their functionalities. In this study, multivariate analysis and g-ogram, a G-SIMS-related spectrum analysis method, were applied to the ToF-SIMS data of polyethylene glycol mixed model sample consisting of the two different average molecular weights, 600 u and 2000 u (PEG600 and PEG2000). As a result, the pure component spectra suggesting PEG600, PEG2000 and polydimethylsiloxane contaminant were discriminated by using multivariate curve resolution (MCR) under the condition of reducing the chemical noise such as heavily fragmented ions and contaminants. Moreover, the major peaks of the pure component spectrum suggesting PEG2000 in MCR were the characteristic fragment ions obtained by the g-ogram result for PEG2000. Therefore, it is indicated that the combination of MCR and g-ogram methods is a promising tool for evaluating the functionalities of the high-functional plastics. Copyright © 2014 John Wiley & Sons, Ltd.