Raman spectroscopy was used to evaluate rank for a series of coals collected adjacent to an igneous intrusion. For these intruded coals, Raman parameters for collotelinite and its coked equivalent show complex changes during the transformation from low to very high levels of maturity (with vitrinite reflectance increasing from 0.55% to 5.0%). With increasing coal rank up to anthracite-level reflectance, the Raman spectra show a decrease in both G band full width at half maximum (FWHM) and D band intensity. The D band shape becomes increasingly asymmetric during the bituminization and de-bituminization stages, whereas in the anthracitization stage, the width of the D and G bands continues to decrease while D band intensity increases. In meta-anthracites (as seen in the ancillary Hunan samples), the G band has a low intensity and undergoes an apparent broadening due to the presence of the D2 band that occurs as a shoulder on the G band. Differences in Raman spectral parameters for vitrinite and inertinite macerals suggest differences in molecular structure between the macerals; as such, the heterogeneous nature of coal may affect Raman spectral results significantly if different maceral types are not considered. Within individual samples, Raman spectra indicate a high level of structural homogeneity between and within vitrinite particles, allowing assessment of differences between samples of different ranks. For vitrinite, the Raman spectral parameters G FWHM and DAs/GA (Area(1100-s)/Area(s-1650), s is the saddle point between 1100 cm−1 and 1650 cm−1) are highly correlated (R2 > 0.9) with vitrinite reflectance (Rr) and maximum alteration temperature (Tpeak), and allow establishment of geothermometers over the temperature range of ~70–300 °C that can be used to evaluate temperatures attained in intruded coals. •Rr increases from 0.55%–5.0% for a coal transect adjacent to an intrusion.•With increased Rr, Raman spectra parameters G FWHM and D band intensity decrease.•G FWHM and DAs/GA are correlated (R2 > 0.9) with Rr and Tpeak.•Two equations were developed to estimate maximum temperature from Raman data.