Rapid assessment of wheat straw at different maturity stages will help to reveal its growth mature and enable better process control that will optimize the sustainable value-added usage. This study explored the potential of near-infrared reflectance (NIR) spectroscopy to quantitatively and qualitatively analyze the multiple chemical composition and calorific value of wheat straw at different maturity stages. Partial least-squares (PLS) and genetic algorithm and partial least-squares (GA-PLS) models were used for NIR spectroscopy analysis. Results showed that PLS models and GA-PLS models could be both used for the estimation of chemical composition and calorific value of wheat straw at three maturity stages, and the GA-PLS method reduced the spectral variables for modeling and provided sensitive spectral variables correlated well with chemical composition and calorific value. NIR spectroscopy could successfully detect the contents of water-soluble carbohydrates (WSC), crude protein (CP), acid detergent fiber (ADF), neutral detergent fiber (NDF), ash, and nitrogen (N). It was also able to quantify dry matter (DM), cellulose (Cel), moisture (Moist), volatile matter (VM), higher heating value (HHV), and lower heating value (LHV). The NIR models for hemicellulose (Hem), lignin (Lig), carbon (C), sulfur (S), and hydrogen (H) had moderate accuracy, and could be used for qualitative analysis, whereas for fixed carbon (FC), it was suitable for screening purposes.