Shoe soles leave trace evidence that can easily be extracted at crime scenes. The Fourier transform infrared spectroscopy, combined with statistical analysis, is highly significant for testing and identifying physical evidence from the sole. In this study, five sole samples that are primarily made of cis-butadiene rubber are selected for identification. The original spectra of the soles that the main materials are the same are identical and difficult to effectively distinguish; therefore, the fingerprint region spectra in the approximate range of 1800?650 cm^-1 are selected for the second-derivative analysis and two-dimensional correlation spectral analysis. The results show that the differences in the spectra of the second derivatives are not significant between samples, whereas the number, position, and response order of the cross peaks in the two-dimensional correlation spectra differ; this can be used to distinguish different sole samples. Therefore, using two-dimensional correlation infrared spectroscopy, shoe soles are useful for identifying physical evidence. Thus, two-dimensional correlation spectroscopy has great potential for application in the field of forensic science.