Coating distinction was easily achieved in many cases because of the presence of dominant ions derived from the components of the coating, which could be observed in the SIMS spectra. In other instances, coating distinction was difficult within a product line because of spectral complexity; for this reason and because of the large numbers of spectra generated in this study, multivariate statistical techniques were used, which allowed the meaningful classification and comparison of spectra. Partial Least Squares (PLS) and Principal Component Analysis (PCA) were applied to quadrupole SIMS data. PCA showed distinct spectral differences between most spectral groups and also emphasized the reproducibility of the SIMS spectra. When using PLS analysis, reasonably accurate coating identification was achieved with the data. Overall, the PLS model is more than 90 percent effective in identifying the spectrum of a particular coating, and nearly 100 percent effective at telling which coating components represented in the PLS models are not present in a spectrum. The level of spectral variation caused by sample bombardment in the SIMS analysis was investigated using Fourier transform infrared spectroscopy (FT-IR) and quadrupole static SIMS. Changes in the FT-IR spectra were observed and were most likely a result of a number of factors that involved the static SIMS analysis; however, the bulk of the sample is unaltered and may be used for further testing. 3 tables, 8 figures, and 44 references