In order to validate the usefulness of 1,5-anhydroglucitol (AG) as an index of glycemic variability, we examined the correlation between the 1,5-AG value and the amplitude of glycemic fluctuation obtained from a continuous blood glucose monitoring system (CGM) in diabetic patients, the difference in the distribution of HbA1c and 1,5-AG among the different therapies, and the changes in the HbA1c and 1,5-AG levels before and after the administration of DPP-4 inhibitor. As a result, a significant negative relationship was observed between 1,5-AG and the standard deviation (SD) of blood glucose and between 1,5-AG and the mean amplitude of glycemic excursions (MAGE). Among the diabetic patients with 1,5-AG<10 μg/ml, the percentage of patients treated with sulfonylurea or insulin was higher than with other therapies, regardless of the HbA1c value, thus suggesting that sulfonylurea or insulin therapy could not sufficiently reduce glycemic variability. The administration of DPP-4 inhibitor decreased the HbA1c values and increased the 1,5AG values, thus suggesting that this medicine could effectively reduce glycemic variability. In conclusion, 1,5-AG is a useful and valuable index of glycemic variability and in order to suppress glycemic variability, it is desirable to adopt a treatment strategy aimed at 1,5-AG≥10 μg/ml.