Summary Background & aims Flavonoids may have cardioprotective effects, but epidemiological evidence on the relationship of dietary flavonoids with diabetes has not been systematically assessed. To examine the association between dietary flavonoids and type 2 diabetes, we performed a meta-analysis on this topic. Methods We searched PubMed through March 2013 for relevant cohort studies that assessed total flavonoids and type 2 diabetes risks. A fixed-effect model was used to calculate the summary risk estimates. Results Four articles consisting of 6 prospective cohorts that involved 18,146 cases and 284,806 participants were identified. The summary relative risk (RR) of type 2 diabetes for the highest intake of total flavonoids compared with the lowest was 0.91 (95% confidence interval (CI): 0.87–0.96). Furthermore, an increase in the total flavonoids intake of 500 mg/d was associated with a significant risk reduction of 5% (RR = 0.95, 95% CI: 0.91–0.98). In subgroup analyses, the observed beneficial effects were observed in US population, in those mean age > 40 years old people and in studies ≥20 years in duration. Conclusions The present meta-analysis indicates that consumption of dietary total flavonoids is associated with a reduced risk of type 2 diabetes.