Near-infrared spectroscopy (NIRS) is a technique that shows many possibilities in the field of identifying chemicals and materials. Plastic food packaging is not excluded of this practical application. In the current recycling system, a plastic identification must be done to avoid the hidden danger caused by defective goods. In an attempt to develop an alternative plastic identification technology, four plastics are collected including Polypropylene (PP), Polyethylene (PE), Polyethylene Terephthalate (PET), and Polystyrene (PS), and a qualitative model with Factor Analysis method will be built. Performance of the model varies substantially if wave bands and multivariate data analysis techniques change. When the wavelength ranges from 4700 to 10000cm-1, plastic food packaging whose spectrum data are optimized by first derivative + vector normalization with 5 factors can be successfully identified, with which a classification error of less than 4% can be reached within the validation set of 100 spectrum. The results obtained indicate that the method proposed in this paper has a real potential for future plastic classification uses.