Objective To analyze the correlation of glycosylated albumin (GA) and glycosylated hemoglobin A1c (HbA1c) with mean amplitude of glycemic excursions (MAGE) and standard difference of blood glucose (SDBG) in elderly type 2 diabetic patients, so as to provide theoretical evidences for complete assessment of GA and HbA1c. Methods A total of 660 elderly type 2 diabetic patients hospitalized in our department from February 2013 to February 2014 were enrolled in this study. Fasting plasma glucose (FPG), HbA1c, GA, fasting C peptide (FCP), 2-hour postprandial blood glucose (2hPBG), and 2-hour C peptide (2hCP) were examined in all the subjects. The 3-day continuous blood glucose monitoring of the enrolled subjects were performed by continuous glucose monitoring system. MAGE and SDBG were used to assess to glycemic excursions. Mean blood glucose (MBG) was used to assess the overall glucose level. Results (1) Among the 660 enrolled subjects, there were 346 men and 314 women with age of (66±6) years, diabetes course of 8.0(4.0?13.8)years, HbA1c of 8.2% (6.7%?9.6%), GA of 21.1% (17.6%?25.9%), MAGE of 5.3(3.9?7.2)mmol/L, SDBG of 2.1(1.6?2.7)mmol/L, and MBG of 8.6(7.3?10.1)mmol/L. (2) Univariate correlation analysis showed that GA was significantly correlated with HbA1c (r＝0.836, P＜0.01), and also positively correlated with FPG, 2hPBG and MBG (r＝0.604, 0.670 and 0.650, respectively, P＜0.01). HbA1c also had positive correlation with FPG, 2hPBG and MBG (r＝0.603, 0.634 and 0.661, respectively, P＜0.01). (3) Univariate correlation analysis indicated that GA had positive correlation with MAGE and SDBG (r＝0.485 and 0.529, P＜0.01). HbA1c was also positively correlated with MAGE and SDBG (r＝0.417 and 0.495, P＜0.01). (4) Stepwise multivariate regression analysis demonstrated that GA level was independently correlated with the glycemic excursion indices, MAGE and SDBG. Conclusion Compared with HbA1c, GA can reflect better postprandial glucose level and glycemic excursions.