Objective Although few studies reported clinical factors associated with glycemic variability in type 2 diabetes, factors associated with absolute and relative measures of glycemic variability in type 2 diabetes without insulin therapy has not been determined by continuous glucose monitoring (CGM) and concurrent measure of fasting C-peptide levels. Methods We analyzed CGM data for subjects with type 2 diabetes who have not been treated with insulin (n = 231) and underwent CGM between Oct. 2009 and Sep. 2011 at Samsung Medical Center. Correlations between clinical factors and both standard deviation (SD) and coefficient of variance (CV) in CGM were analyzed by multiple regression. Results CV, but not SD, inversely correlated with the minimum glucose level during CGM. In univariate analysis, age, duration of diabetes, and HDL levels correlated with SD and CV. HbA1c correlated with both SD, and LDL levels inversely correlated with CV. Higher SD (45.5 ± 17.7 vs. 32.5 ± 16.0, P < 0.001) and CV (28.3 ± 8.5 vs. 23.0 ± 8.1, P < 0.001) were observed in subjects treated with sulfonylurea. In multivariate analysis, age, BMI, HbA1c, HDL and triglyceride levels, and use of sulfonylurea correlated with SD. HDL and triglyceride levels, and use of sulfonylurea correlated with CV, and LDL levels inversely correlated with CV. Conclusion CV was determined by factors different from those affecting SD. Besides the use of sulfonylurea, some of these factors were indicators of higher insulin sensitivity such as higher HDL and lower LDL levels.