In this paper, near infrared spectroscopy combined with a wavelength selection method was used to identify fermented vinegar and blended vinegar. Competitive adaptive reweighted sampling (CARS) method was employed to select the key wavelengths. For classification of different vinegars, 11 wavelengths were selected and a PLS-DA model was calculated based on these selected wavelengths. The average error rate (AER) obtained by repeated double cross validation (RPCV) was 6.95%. The results indicated that NIR spectroscopy combined with a wavelength selection method could be applied as a fast alternative method for the identification of different vinegars.