Tooth is of great significance to human health. With age, the characteristics of the teeth will change. At present, different detection methods have been developed to detect the characteristics of teeth. However, the existing detection methods have shortcomings. In view of the effective characterization of teeth characteristics in different age groups, this paper aims to explore spectral polarization method, namely a non-destructive and low-loss detection method, which is a useful supplement to conventional detection methods. The method for spectral polarization to effectively characterize the tooth characteristics of different ages was proposed. Tooth samples from 7 different age groups, such as 10-20 years old, 20-30 years old, 30-40 years old, 40-50 years old, 50-60 years old, 60-70 years old and over 70 years old, were selected; 4 different observation spectrum bands such as 450nm, 550nm, 670nm, 870nm were selected; the polarization parameters were selected to describe the spectral polarization characteristics of the teeth, and a polynomial correlation mathematical model was constructed. The experimental results showed that the tooth samples in the same age group showed a negative correlation between spectrum band and polarization characteristics. The polarization characteristics of tooth samples in subjects aged 50-60 years old reached the peak value for the same observation band. Construction of the model could effectively interpret the coupling correlation between tooth samples and spectral polarization characteristics in different age groups. The research content of this paper effectively expands the method of tooth characteristic detection, reveals that spectral polarization can effectively characterize tooth characteristics, and develops a novel non-destructive and lowloss polarization spectral detection technology.