A rapid label-free approach for molecular histopathology is presented and reviewed. Broadband vibrational spectra are generated by nonlinear interferometric vibrational imaging (NIVI), a coherent anti-Stokes Raman scattering (CARS)-based technique that uses interferometry and signal processing approaches to acquire Raman-like profiles with suppression of the non-resonant background. This allows for the generation of images that provide contrast based on quantitative chemical composition with high spatial and spectral resolution. Algorithms are demonstrated for reducing the diagnostic spectral information into color-coded composite images for the rapid identification of chemical constituents in skin, as well as differentiating normal from abnormal tissue in a pre-clinical tumor model for human breast cancer. This technology and methodology could result in an alternative method to the traditional histological staining and subjective interpretation procedure currently used in the diagnosis of disease, and has the potential for future in vivo molecular histopathology.