The aim of this study was to generate calibration equations to predict the nutritional chemical composition of the Italian rye grass (RG) (Lolium multiflorum Lam) by near infrared spectroscopy (NIRS). A total of 75 samples of RG of different harvesting weeks were collected from the IVITA Research Center in Huancayo (Peru). Spectrum capture was performed using NIRS and the chemical analysis was done for reference of the following components: crude protein (CP), ether extract (EE), total ash (CZ), crude fibre (CF) and neutral detergent fibre (NDF). A calibration and validation model by partial least squares (PLS) was developed and the correlation coefficient (R), coefficient of determination (R2), root mean square error of calibration (RMSEC), root mean square error of prediction (RMSEP), ratio range with error (RER) and residual predictive deviation (RPD) were used as statistics of accuracy and precision. Proximate analysis means were: PC = 19.02%, EE = 4.53%, CZ = 12.79%, FC = 16.50% and NDF 60.98%. High values of R2 and low values of RMSEC and RMSEP were obtained for PC (0.96, 1.02, 1.19), EE (0.94, 0.29, 1.05), CZ (0.90, 0.57, 0.92) and NDF (0.90, 1.01, 1.25, respectively). The largest RER (22.34) and RPD (4.90) were obtained for EE. It is concluded that the calibration and validation equations obtained by NIRS enable optimal quantitative prediction of PC, EE, CZ and NDF in Italian rye grass (Lolium multiflorum Lam).