Fourier transform infrared (FTIR) chemical imaging is a strongly emerging technology that is being increasingly applied to examine tissues in a high-throughput manner. The resulting data quality and quantity have permitted several groups to provide evidence for applicability to cancer pathology. It is critical to understand, however, that an integrated approach with optimal data acquisition, classification, and validation is necessary to realize practical protocols that can be translated to the clinic. Here, we first review the development of technology relevant to clinical translation of FTIR imaging for cancer pathology. The role of each component in this approach is discussed separately by quantitative analysis of the effects of changing parameters on the classification results. We focus on the histology of prostate tissue to illustrate factors in developing a practical protocol for automated histopathology. Next, we demonstrate how these protocols can be used to analyze the effect of experimental parameters on prediction accuracy by analyzing the effects of varying spatial resolution, spectral resolution, and signal to noise ratio. Classification accuracy is shown to depend on the signal to noise ratio of recorded data, while depending only weakly on spectral resolution.