Abstract: To evaluate the quality consistency of tobacco flavors, a full-time multi-wavelength synthetic HPLC fingerprint analysis method was established by combining five wavelengths (240, 254, 278, 330, and 360 nm) simultaneous detection with chromatographic peak screening at each wavelength. Through comparison with four different methods, including the angle cosine method, an algorithm suitable for evaluating the similarity of tobacco flavor fingerprints was obtained and discussed, and the similarity thresholds for determining the quality consistency of 31 tobacco flavors were determined based on sensory evaluation. The results showed that: 1) The full-time multi-wavelength synthetic fingerprints could fully reflect the intrinsic quality information of tobacco flavors. 2) The modified Nei coefficient algorithm was more suitable for evaluating the similarity of tobacco flavors. 3) The determined similarity thresholds could be used to separately evaluate the quality consistency of 31 tobacco flavors. In conclusion, this method was accurate and could be used for quality control of tobacco flavors.