In order to evaluate the relationship between clinical markers of glycemia and glucose excursion, we performed 48-hour continuous glucose monitoring (CGM) in 43 diabetic patients. For the clinical markers, HbA(1c), glycoalbumin (GA), and 1,5-anhydroglucitol (1,5-AG) were measured, and for the parameters of glucose excursion from CGM, average glucose (AG), standard deviation of glucose (SD), the area under the curve for glucose levels >180 mg/dL (AUC(>180)), and the difference between the maximum and minimum glucose levels during 48 hours (DeltaG(48hr)) were analyzed. All patients were treated without any changes of the dosages of oral anti-diabetic agents or insulin for at least the previous 3 months with coefficient of variation (CV) of HbA(1c) less than 4 %. In results, while HbA(1c) did not show any single correlation with AG, SD, AUC(>180), or DeltaG(48hr), both GA and 1,5-AG were significantly related to all these parameters. Furthermore, GA significantly correlated to all CGM parameters, and SD significantly correlated to GA in multiple regression analyses. These results suggest that GA may be a different marker from HbA(1c) for diabetic complications, because GA, but not HbA(1c), may reflect not only short-term average glucose but also fluctuation of glucose.