This meta-analysis aimed to assess the gender-specific differences in the relationship between circulating leptin levels and risk of type 2 diabetes. Published prospective studies that reported the association of leptin levels with risk of type 2 diabetes for a certain gender or those that reported gender-specific associations were considered. Dose-response relationships were assessed by the generalized least squares trend estimation and summary relative risks (RRs) with 95% confidence interval (CI) were computed with the random-effects model. Stratified and sensitivity analyses were also performed to investigate potential sources of heterogeneity. Overall, 11 prospective studies were identified. The summary RR for an increment in leptin levels of 1-log ng mL−1 was 1.37 (95% CI, 1.13–1.66) for men and 0.96 (95% CI, 0.90–1.03) for women. The differences between genders were statistically significant (P for interaction = 0.006). Subgroup and sensitivity analyses generally confirmed the robustness of these findings. Furthermore, the increased risk in men appeared non-linear, with a tendency to plateau at high levels (P for non-linearity = 0.03). Little evidence of publication bias was found. Collectively, higher leptin levels were found to be associated with elevated risk of type 2 diabetes in men but not in women.