The feasibility of noninvasive visible and near infrared (VIS-NIR) spectroscopy for determining vintage year of Chinese rice wine was investigated. VIS-NIR spectra of 100 samples were collected in transmission mode in 600-mL square brown glass bottles by a fiber spectrometer system. Discriminant models were developed based on discriminant analysis (DA) together with raw, 1st, and 2nd derivative spectra. The concentration of alcohol content, total acid, and degrees Brix was determined to validate the VIS-NIR results. The calibration result for raw spectra was better than that for 1st and 2nd derivative spectra. The percentage of samples correctly classified for raw spectra was 97.1%. For the sample groups in the vintage years of 2002, 2003, and 2004, the samples were all correctly classified. And for the 2000 and 2001 sample groups, the percentage of samples correctly classified was 92.9%. In validation analysis, the samples were all correctly classified. The results demonstrated that the VIS-NIR spectroscopic technique could be used as a noninvasive and rapid method for predicting vintage year of bottled Chinese rice wine.