A systematic investigation, based on highly spectrally resolved Raman spectroscopy, was undertaken to research the efficacy of vibrational assessments in locating chemical and crystallographic fingerprints for the characterization of dental caries and the early detection of non-cavitated carious lesions. Raman results published by other authors have indicated possible approaches for this method. However, they conspicuously lacked physical insight at the molecular scale and, thus, the rigor necessary to prove the efficacy of this spectroscopy method. After solving basic physical challenges in a companion paper, we apply them here in the form of newly developed Raman algorithms for practical dental research. Relevant differences in mineral crystallite (average) orientation and texture distribution were revealed for diseased enamel at different stages compared with healthy mineralized enamel. Clear spectroscopy features could be directly translated in terms of a rigorous and quantitative classification of crystallography and chemical characteristics of diseased enamel structures. The Raman procedure enabled us to trace back otherwise invisible characteristics in early caries, in the translucent zone (i.e., the advancing front of the disease) and in the body of lesion of cavitated caries.