Shock pressures in framework silicates and glasses of shocked rocks are deduced from the values of their refractive index (n); n is directly determined in a polished thin section by measuring reflectivity using a microreflectometry technique adapted to transparent minerals. The technique is tested on experimentally and naturally shocked crystalline rocks and on unshocked material. The technique is fast, nondestructive, directly adapted to the study of naturally shocked materials and can be used for both crystalline and glassy phases whatever the structure and the composition. The major limitation actually comes from the scarcity of experimental data relating shock pressure and refractive index for calibration. Accuracy is excellent for tectosilicates in the range 15--45 GPa. The small size of the measured field (10 μm or less) allows the recognition and quantitative estimates of pressure gradients within shocked materials. All intermediate values for n are found ranging from n of the unshocked minerals to n of the corresponding diaplectic glass. Some diaplectic glasses of quartz and plagioclase An27 have considerably higher refractive index than previously known,1.525 and 1.538 respectively, suggesting that diaplectic glasses of quartz and plagioclase can form at relatively lower pressures than previously thought, about 23.5 and 27.5 GPa, respectively. The ranges of n of quartz and diaplectic glass of quartz are distinct and do not overlap in the experimentally shocked grains. The same feature characterizes An27, while the ranges widely overlap in microcline of a naturally shocked granite. Diaplectic crystals and diaplectic glasses are possibly a mixture of crystalline and amorphous submicroscopic domains. Diaplectic glass of microcline and oligoclase gives pressure estimates 15--20% higher than does diaplectic glass of quartz and feldspars of a gneiss sample shocked at 35.3 GPa in the same naturally and artifically shocked rocks. Average pressure measured in quartz and are significantly lower than the shock pressure of the whole rock. Further work is required to understand these discrepancies. Possible applications of the technique are introduced.