The anaerobic digestibility of various biomass feedstocks in biogas plants is determined with biochemical methane potential (BMP) assays. However, experimental BMP analysis is time-consuming, costly and challenging to optimise stock management and feeding to achieve improved biogas production. The aim of the present study is to develop a fast and reliable model based on near-infrared reflectance spectroscopy (NIRS) for the BMP prediction of urban organic waste (UOW). The model comprised 87 UOW samples. Additionally, 88 plant biomass samples were included, to develop a combined model predicting BMP. The coefficient of determination (R2) and root mean square error in prediction (RMSEP) of the UOW model were 0.88 and 44 mL CH4/g VS, while the combined model was 0.89 and 50 mL CH4/g VS. Improved model performance was obtained for the two individual models compared to the combined version. The BMP prediction with NIRS was satisfactory and moderately successful.