Abstract: On the basis of existing technology, the secondary development of RIPP`s rapid crude oil evaluation system by near infrared was conducted. A rapid evaluation model was established based on the database of crude oils processed in refineries by combination of crude data library spectrum fitting and identify technique. The secondary developed rapid evaluation system was applied to A, B two refineries to identify and predict the kind and properties of the crude. The results demonstrate that the established evaluation system can identify well the crude oils processed by refineries under the conditions of the perfect database. The density of crude oil, acid value, content of sulfur and carbon residue of the two refineries can be predicted with small deviation from the standard (RMSEP). The repeatability is equal to the standard method.