Background: Evidence of increased type 2 diabetes (T2D) risk associated with potatoes consumption is equivocal. We aimed to perform a meta-analyses on the association between potatoes consumption and T2D risk in prospective cohort studies. Methods: Studies published prior to 31 Aug 2016 were identified in PubMed, EMBASE, and Web of Science. Pooled relative risks (RR) and 95% confidence intervals (95%CI) based upon the highest vs. lowest category of potatoes consumption in each study were calculated in meta-analysis using random-effects models. Dose-response meta-analysis was fitted using generalized least squares regression in order to quantify the association between potatoes consumption and T2D risk. Results: The pooled RR comparing the highest vs. lowest category of potato consumption was 1.077 (95%CI: 1.005, 1.155). Dose-response meta-analysis revealed T2D risk increased 3.5% (RR=1.035, 95% CI: 1.004-1.067) for additional three serving per week serving of potato. The pooled RR comparing the highest vs. lowest category of French fries consumption was 1.362 (95%CI: 1.004, 1.850). Dose-response meta-analysis indicated T2D risk increased 18.7% (RR = 1.187, 95% CI: 1.067-1.321) for additional three serving per week of French fries. Conclusion: This meta-analysis support a significant positive association between high potatoes consumption and risk of T2D, especially the consumption of French fries.