FT-IR microspectroscopic maps of unstained thin sections from human melanoma and colon carcinoma tissues were obtained on a conventional infrared microscope equipped with an automatic x, y stage. Mapped infrared data were analyzed by different image re-assembling techniques, namely functional group mapping ("chemical mapping") and, for the first time by cluster analysis, principal component analysis and artificial neural networks. The output values of the different classifiers were recombined with the original spatial information to construct IR-images whose color or gray tones were based on the spatial distribution of individual spectral patterns. While the functional group mapping technique could not reliably differentiate between the different tissue regions, the approach based on pattern recognition yielded images with a high contrast that confirmed standard histopathological techniques. The new technique turned out to be particularly helpful to improve discrimination between different types of tissue structures in general, and to increase image contrast between normal and cancerous regions of a given tissue sample.