Techniques previously utilized for chemical analysis of enamel surfaces are generally unsuitable for in-vivo testing. This study explored the utility of internal reflection spectroscopy (IRS) as a non-destructive method for enamel-surface analysis.Powders (enamel plus reference inorganic materials) and solids (flattened and unaltered enamel slabs) were analysed by IRS. Most samples were mounted in a Wilks Model 45 Micro-ATR Accessory and scanned from 2 to 25 pm to obtain i.r. spectra. Comparability of IRS data with that of transmission methods was established. Typical apatitic ~a phosphorous-oxygen absorption was observed at 1090 cm-' and 1040 cm-' in spectra of natural fluorapatites and enamel powders. Spectral differences were demonstrated between powdered and intact enamel. The presence of carbonate was conhrmed in both powdered and solid enamel. Intact surface enamel and fluorapatites failed to absorb at 630 cm-'. Weak absorption at 630 cm-r after disking of the enamel surface was interpreted as a relative loss of fluorapatite due to mechanical reduction to subsurface levels.Internal reflection spectroscopy enabled subtle chemical differences in enamel surface to be distinguished. This non-destructive method offers promise for in-vivu enamel surface analysis. SURFACE enamel represents the initial barrier to dental caries. Investigators have long realized that a thorough knowledge of the chemical nature of enamel is a prerequisite to the effective control of dental caries at its inception. While this rationale for enamel research is unquestioned, inability to detect subtle changes in the actual enamel surface is a shortcoming of available techniques. In addition, most methods require sample destruction to accomplish the analytical procedure. Thus available testing is limited not only by insensitivity to surface properties but also by unsuitability for in-viz90 analysis.At present, the dynamics of the incipient carious lesion are difficult to assess with precision. Magnification of tooth surfaces to several hundred times original dimensions is possible with light microscopy. Scanning electron microscopy affords even greater resolution (HOFFMAN, MCEWAN and DREW, 1968). However, the chemical nature of surface changes cannot be determined by these methods. In addition to the * Present address: