In-line monitoring of chemical processes is desired for its numerous advantages, such as lesser waste, lower developmental cycle time, and lesser costs. In this study, a methodology is presented for estimating polymer rheological properties using fiber-optic near-infrared (NIR) spectroscopy. Predictive calibration models are developed for simultaneous, in-line monitoring of polymer melt flow index (MI) and comonomer concentration for a system of poly( ethylene vinyl acetate) copolymers. The NIR spectra of flowing, molten poly( ethylene vinyl acetate) (EVA) copolymers are collected in a flow cell attached to a single-screw extruder. Multivariate statistical regression analysis is presented for correlation of MI and absorbance in the methylene (C--H) stretch, first overtone NIR wavelength region (1620-1840 nm). Results for simultaneous, real-time monitoring of MI and comonomer composition are discussed in detail.