Objective: This study focuses on designing a more accurate identification model of lettuce storage time. Methods: The near-infrared spectral data from extraction feature of the preprocessed lettuce was obtained by principal component analysis (PCA), discriminant principal component analysis (DPCA) and fuzzy discriminant principal component analysis (FDPCA) respectively. An algorithm of higher accuracy in storage time discrimination was explored, and then a lettuce storage time discriminant model based on FDPCA was established. Results: The identification accuracy raised dramatically after FDPCA was used to extract feature. When employing PCA, DPCA and FDPCA algorithms, the highest accuracies achieved were 46.67%, 86.67% and 93.33% respectively. Conclusion: This discrimination model of employing near-infrared spectroscopy and FDPCA was characterized by high accuracy and superiority.