In this work, the physical and electromagnetic backgrounds of near infrared spectroscopy including operation of chemometrics, the processing of spectral data and assessment of the chemometric models were discussed regarding quality and outliers. In addition, the current state of the art has been explained. The second part of the research is the determination of components and material properties by correlation of spectral data. Various additives which are used in flame retardant systems have been investigated. The determination was based on the state of the art, but was extended, regarding the quantity of different materials. Additionally PVC nanocomposites were examined regarding their composition. The chemometric models of all additives studied showed very good quality indicators whereby a quantitative characterization was possible in all cases. Different mechanical properties such as Young’s modulus and tensile strength but also structural properties such as interparticle distance or the layer spacing on the example of layered silicate based polymer nanocomposites, could also be quantitatively characterized. In addition a characterization of the rheological parameters storage and loss modulus has been tried. With some restrictions, it was also possible to create a partly useful correlation. Also a chemometric model was created to determine extensional rheological properties, represented by the drawing force. This force was measured online with the help of a bypass directly during production. A quantitative determination could be realized. The evidence that polymer nanocomposites can be characterized in terms of their mechanical as well as structural properties and thus a quality control is feasible, is completely new regarding the extend of the investigations. The third part of the work deals with the feasibility analysis of different relevant issues in polymer processing. Thus for example it was possible to connect batch difference, occurring during profile extrusion of PVC with differences in spectral data, although the conventional inspection of incoming goods declared them to be in order. In addition it could be shown that even the slightest contamination during the extrusion process can be made visible. Furthermore, it was shown that the chain scission of PET during processing as well as the moisture absorption of hygroscopic polymers can be characterized by significant changes in the spectra. Additionally the use of NIR spectroscopy for the determination of the residence time distribution in different extrusion processes was investigated. An approach could be developed which allows the end user to investigate the residence time distribution at all extrusion processes with a big variety of different polymers without the need of chemometric modeling. This method is completely new. This work has clearly demonstrated how flexible NIR spectroscopy can be used and how powerful this method can be in terms of an efficient quality control in polymer processing industry and which perspective this method has in future.