In order to rapidly and accurately discriminate whether fishmeal was adulterated with meat and bone meal (MBM) or not, representative fishmeal and MBM samples were collected in China. 201 calibration samples were prepared, including 111 samples adulterated with MBM (1%-33%) and 90 pure fishmeal samples. 113 external validation samples were prepared independently, including 74 samples adulterated with MBM (1%-33%) and 39 pure fishmeal samples. The spectra were scanned from 400 nm to 2 498 nm. The effects of spectrum pretreatment methods and spectrum region on the calibration results were considered. The calibration equations were established by DPLS. A DPLS discriminant analysis model was developed based on spectra region from 400 nm to 2 498 nm with mathematic pretreatment 2-8-6-1 and scatter correction treatment the standard normal variate (SNV). The accurate discriminant rate of calibration and external validation was 95.7% and 95.6% respectively. For the adulterated samples (MBM mass percentage≥5%), the accurate discriminant rate was 100%. The results of the study indicate that NIRS can provide a rapid discriminination method for meat and bone meal in fishmeal.