OBJECTIVES: Sustained-release glipizide has a more appealing pharmacological profile over immediate-release glipizides. However, individual trials have not reliably ascertained its effects. This study systematically reviewed the trials that compared the effects of sustained-release glipizide with the conventional immediate-release glipizide for type 2 diabetes mellitus. METHODS: We searched Medline, EMBASE, the Cochrane Library and three other Chinese databases from their inception to July 2008, as well as screened the reference lists of eligible trials and reviews, and contacted the company (Pfizer) for unpublished data. Two reviewers judged the trial eligibility, assessed the validity, and extracted data independently. We pooled the trial data using the random-effect model and explored the heterogeneity by the pre-specified variables. RESULTS: A total of 16 trials (n=1033) were included. Sustained-release glipizide significantly decreased FPG by 0.33mmol/L (weighted mean difference, 95%CI 0.05 to 0.61), postprandial insulin levels by 3.18 ìIU/ml (0.89 to 5.47), and C-peptide by 0.12 ng/ml (0.04 to 0.20). Sustained-release glipizide did not reduced the HbA1c (-0.02, -0.20 to 0.15), postprandial plasma glucose (0.38, -0.47 to 1.22), fasting insulin levels (1.20, -0.14 to 2.54). No statistical differences were found in the change of total cholesterol (0.09, -0.06 to 0.23), triglyceride (0.13, -0.04 to 0.29), LDL (-0.03, -0.12 to 0.05), HDL(0.04, -0.02 to 0.10), and hypoglycemia (RR 0.79, 95%CI 0.22 to 2.86). No trials reported diabetes-related morbidity and mortality. CONCLUSIONS: Sustained-release glipizide could reduce FPG, postprandial insulin levels, and C-peptide, but has not shown benefits in reducing HbA1c, PPG, and fasting insulin levels when compared to immediate-release glipizide. Uncertainty remained in the benefits of sustained-release glipizide over immediate-release glipizide. This was mainly driven by the small sample size of the trial and lack of long-term morbidity and mortality data.