The hydrolysis of polylactide (PLA) occurs during melt processing, leading to product defects in, for example, appearance and mechanical properties. Hydrolyzed PLA products, whose mechanical properties were deteriorated, must be detected during processing to ensure the quality control of PLA products. In this study, near‐infrared (NIR) hyperspectral imaging was applied for evaluating the extent of hydrolysis. NIR spectra in the wavelength range from 1200 to 1600 nm were changed by (1) hydrolysis induced by molding and (2) crystallization induced by annealing. The changes in the NIR spectra mediated by these two factors were distinctly different but overlapping, leading to difficulty in evaluating PLA at only a single wavelength. Partial least squares regression was introduced for detecting the hydrolyzed PLA. NIR imaging combined with the constructed partial least square models clearly visualized the differences in the extent of hydrolysis in hydrolyzed PLA under varying degrees of crystallization. © 2017 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2018, 135, 45898.