Abstract: Abstract: Freshness is one of the most important attributes to assess meat quality, sale ability, and consumption. However, traditional freshness detection cannot fully meet the rapid and non-destructive monitoring of meat at present, due to the labor intensity and destructive measurement. Therefore, a rapid and non-destructive fluorescence imaging technique can be expected to evaluate lamb meat freshness in the meat industry. The fluorescence substances in meat can be identified as a potential indicator of freshness. Accordingly, this study aims to clarify the relationship between fluorescent substances and freshness index in meat. The fluorescent characteristic markers were also screened to characterize the change in lamb meat freshness. The longissimus dorsi muscle of sheep at 24 h postmortem was prepared in the air-sealed packaging, and then stored at 0℃ for 0, 1, 3, 5, 7, 9, 11, 13, 15, 17, and 21 days. The specific contents were measured in the lamb meat, including the tyrosine (Tyr), adenosine triphosphate (ATP), nicotinamide adenine dicylate (NADH), tryptophan (Trp), riboflavin (VB2), protoporphyrin (PPIX), and several freshness indexes (TVB-N, TBARS, TVC, pH value, and meat color). The results showed that the TBARS, TVB-N, and TVC of the lamb meat increased significantly during storage, whereas, the pH value decreased first and then increased, and the L*, a*, and b* values increased first and then decreased in the meat. The potential fluorescent markers (Tyr, Trp and PPIX) of freshness showed a significant increasing trend, while the NADH and ATP were the opposite change pattern, and the VB2 was a trend of increase firstly and then decrease in the meat. The correlation and cluster analysis demonstrated that the ATP, NADH, Tyr, Trp, and PPIX shared a significant positive correlation with the TVB-N and TVC, where the correlation coefficients ranged from 0.60 to 0.88. The fluorescence features molecule was verified at 4 ℃. It was found that the fluorescence features molecule was consistent with that at 0℃. Stepwise regression analysis was conducted with the ATP, NADH, Tyr, Trp, and PPIX indexes. Among them, the TVB-N (mg/100g) =87.926X2-47.222X3-26.773 (R2=0.926) and TBARS (mg/kg) =1.850X2-2.227X3-1.072 (R2=0.878) were obtained for the optimal regression models to predict the TVB-N and TBARS in the lamb meat by Tyr and Trp, respectively. The optimal regression model of the TVC was the TVC (lg CFU/g)=-0.018X1+15.131X2-2.273 (R2=0.952) by the NADH and Tyr. Thus, the NADH, Tyr, and Trp can be expected to serve as fluorescent characteristic markers for the detection of lamb meat freshness during storage. The findings can provide a theoretical basis for the subsequent development of lamb freshness detection using hyperspectral fluorescence imaging.