The autofluorescence spectra of hard dental tissues, both in normal and pathological areas were investigated in this study. The measurements were performed both on the intact hard tissues of the examined teeth, such as enamel, dentine, cementum, and root canal, and on the tissues pathologically affected by caries (superficial, intermediate, and deep). Various laser wavelengths (337 nm, 488 nm, and 514 nm) were used to irradiate the dental surfaces and a computer-controlled spectrograph captured the fluorescent spectra. The emission signals were stored, measured, analyzed and quantified in terms of wavelength distribution and the relative photon intensity. Results indicated that the fluorescent spectra from healthy enamel, dentine, and cementum were almost identical in form, depending on the excitation wavelength. The intact and affected hard tissues were greatly different in the integral fluorescent intensity. Healthy areas were found to produce the most pronounced fluorescent intensity, whereas the carious regions produced the weaker fluorescent intensity. Independently of the laser excitation wavelength, dentin regions were found to produce the most pronounced fluorescent intensity than any other dental component. The fluorescence signal of carious affected dental structure revealed a reed shifted spectral curve, more pronounced after 488 nm excitation. There was a pronounced red shift for deep caries (crown -- root caries), after ultraviolet laser excitation. Excitation with visible wavelengths did not produce such differences between intact and cervical, deep carious affected tissue. Using a monochromatic light source without any light output at the wavelengths of fluorescence, e.g. a laser with the appropriate filters, the difference in fluorescence between intact and carious enamel was generally easy to observe. Finally, we found that the blue line of an argon ion laser is preferable for superficial caries detection, while the ultraviolet emitting nitrogen laser induces better discrimination in deep caries diagnosis.