Glycation albumin(GA) has emerged as a biomarker for diagnosing prediabetes and diabetes, which is shorter than the "gold standard" reflection period of blood glucose detection. However, the GA response is the global glycation rate of albumin, individual amino acid residues might be sensitive indicators of alterations in blood glycation levels. Here we report a method to analyze the glycated peptides that might provide a better understanding of fluctuations of blood glucose concentrations. In this study, a strategy to quantify the depletion of the digestion peptides of albumin for the diagnosis of type 2 diabetes mellitus (T2DM). First of all, a combination of 13C glucose labeling and HPLC-Q-TOF M S/MS were used to screen 16 glucose-sensitive peptides from in vitro HSA and serum models, which represent glucose-sensitive sites on HSA. Furthermore, the candidate peptide biomarkers were validated in 72 clinical samples (44 diabetic patients and 28 healthy controls) using label-free LC-ESI-MRM. Finally , three putative sensitive peptides (VAHRFKDLGEE, FKPLVEEPQNLIKQNCE and NQDSISSKLKE) from the albumin exhibited good specificity and sensitivity based on receiver operating characteristic analysis. The putative peptide biomarkers were verified as the promising biomarkers for the diagnosis and assessmen of diabetes mellitus with clinical patterns.