Some studies have shown Helicobacter pylori (H. pylori) infection to be associated with diabetes mellitus, but the relationship remains controversial. This meta-analysis was designed to quantify the association between H. pylori infection and diabetes. A computerized search of PubMed and Embase was carried out. Studies that provided data on H. pylori infection in both diabetes and control groups were selected. An unconditional logistic regression model was used to analyze potential parameters related to H. pylori prevalence. Subgroup analyses were conducted for types of diabetes, methods of detection, geographical distribution, hemoglobin A1c (HbA1c) levels and evidence grade. Forty-one studies were identified, involving 14,080 patients, with a total H. pylori infection rate of 42.29%. The OR for H. pylori infection was increased to 1.33 (95% CI: 1.08-1.64; P=0.008) among the patients with diabetes. Subgroup analysis revealed a significantly higher infection rate of H. pylori in the type 2 diabetes group versus the control group: OR=1.76, 95% CI: 1.40-2.21, P<0.00001. The pooled data suggests a trend toward more frequent H. pylori infections in diabetes patients, especially in type 2 diabetes patients. As this is a meta-analysis of observational studies, more randomized controlled trials should be done in the future. Copyright © 2012 Elsevier Ireland Ltd. All rights reserved.