Rapid and accurate determination of chemical composition of rice straw is crucial to predictions of its value and nutritive quality. In this work, calibration models using near-infrared reflectance spectroscopy (NIRS) for determination of rice straw analytical parameters such as total ash, insoluble ash, moisture, cellulose, hemicellulose and Klason lignin were studied using a diverse group of rice straws. The NIRS calibration models for chemical compositions of rice straw were derived by partial least-squares (PLS) regression and prediction of chemical composition of independent rice straw samples showed these models to be rapid and accurate, giving R 2 -values higher than 0.85. Such NIRS calibration models are appropriate for the determination of the quality of rice straw samples.