Abstract Christy, A.A., Velapoldi, R.A., Karstang, T.V., Kvalheim, O.M., Sletten, E. and Telnaes, N., 1987. Multivariate calibration of diffuse reflectance infrared spectra of coals as an alternative to rank determination by vitrinite reflectance. Chemometrics and Intelligent Laboratory Systems , 2:199-207. Sixty-four randomly selected, vitrinite-rich coal samples were subjected to petrological, spectrometric and multivariate data analysis. Diffuse reflectance Fourier transform infrared spectra in a Kubelka-Munk format were reduced by a maximum entropy reduction method and were calibrated against vitrinite reflectance using partial least squares regression. Two calibration models were calculated. One model was calculated for coal samples with vitrinite reflectance from 0.38 to 1.08, i.e., including most of the maturity range defined as the “oil window”. Outliers or non-population members were identified by an iterative process, leaving a total of 41 coal samples. The cross-validated model gave an absolute prediction error of ± 0.09, a value of the same order of magnitude as the average standard deviation of the vitrinite reflectance measurements (1 s = 0.06). The second model was calculated for eighteen samples with maturity from 1.32 to 3.54. An absolute prediction error of ±0.15 was obtained, reflecting the higher uncertainty in the vitrinite reflectance measurements for the higher-ranking coals.