Diabetes mellitus is an incurable disease,so it is necessary to establish a model to screen biomarkers for early warning in order to minimize the likelihood of long-term complications.Currently,advanced glycation end products(AGEs)are considered to be biomarkers of many diseases,such as diabetes and its complications.In this study,a model for further proteomics study was established to analyze the glycation of HSA with 18 O-labeling strategy.30 peptides were randomly selected to optimize tryptic digestion and 18 O-labeling condition by HPLCESI/TOF.The best tryptic digestion condition was:HSA∶Trypsin=50∶1,w/w for 20 h.The best 18 O-labeling condition was to dilute urea to 1 M and adjust KH2PO4—K2 HPO4 buffer pH to 6.0to give a final labeling efficiency of 98.5±0.7%.The inter-and intra-day precisions and stability were satisfactory.This model was established and optimized for further quantitative proteomics study. Diabetes mellitus is an incurable disease, so it is necessary to establish a model to screen biomarkers for early warning in order to minimize the likelihood of long-term complications. Current, advanced glycation end products (AGEs) are be bioiskers of many diseases , such as diabetes and its complications.In this study, a model for further proteomics study was established to analyze the glycation of HSA with 18 O-labeling strategy. 30 peptides were randomly selected to optimize tryptic digestion and 18 O-labeling condition by HPLCESI / TOF.The best tryptic digestion condition was: HSA: Trypsin = 50: 1, w / w for 20 h.The best 18 O-labeling condition was to dilute urea to 1 M and adjust KH2PO4-K2 HPO4 buffer pH to 6.0to give a final labeling efficiency of 98.5 ± 0.7%. The inter-and intra-day precisions and stability were satisfactory. This model was established and optimized for further quantitative proteomics study.