In the tire industry, synthetic styrene-butadiene rubber (SBR), butadiene rubber (BR), and isoprene rubber (IR) elastomers are essential for conferring to the product its properties of grip and rolling resistance. Their physical properties depend on their chemical composition, i. e., their microstructure and styrene content, which must be accurately controlled. This paper describes a fast, robust, and highly reproducible near-infrared analytical method for the quantitative determination of the microstructure and styrene content. The quantitative models are calculated with the help of pure spectral profiles estimated from a partial least squares (PLS) regression, using (13)C nuclear magnetic resonance (NMR) as the reference method. This versatile approach allows the models to be applied over a large range of compositions, from a single BR to an SBR-IR blend. The resulting quantitative predictions are independent of the sample path length. As a consequence, the sample preparation is solvent free and simplified with a very fast (five minutes) hot filming step of a bulk polymer piece. No precise thickness control is required. Thus, the operator effect becomes negligible and the method is easily transferable. The root mean square error of prediction, depending on the rubber composition, is between 0.7% and 1.3%. The reproducibility standard error is less than 0.2% in every case.