FTIR spectroscopy is a widely used technique that provides insights into disease processes at the molecular level. Due to its numerous advantages it is becoming an increasingly powerful tool for the study of biological materials and has the potential to become an excellent diagnostic method, especially considering the low cost of transflection substrates. However, questions about the usefulness of the transflection measurement mode due to the complicated nature of physical processes occurring during the measurement and in particular the Electric Field Standing Wave (EFSW) effect have been raised. In this paper we present a comparison of the two most common FT-IR measurement modes: transmission and transfection using healthy and pathologically altered tissue (histiocytic sarcoma). We found that the major differences between normal and cancerous tissue were associated with changes DNA and carbohydrate content. In particular we identified a band at 964 cm(-1) assigned to a nucleic acid phosphodiester backbone mode, which appeared more pronounced in cancerous tissue irrespective of the substrate. We applied Principal Component Analysis, Unsupervised Hierarchical Cluster Analysis and k-means clustering to transmission and transflection substrates and found that both measurement modes were equally capable of discrimination normal form cancerous tissue. Moreover, the differences between spectra from cancerous and normal tissue were significantly more important than the ones arising from the measurement modes.