Objective To evaluate the association between methylenetetrahydrofolate reductase (MTHFR) C677T gene polymorphism and depression risk.Methods Available studies containing genotype frequencies of MTHFR C677T were chosen,and pooled odds ratio (OR) with 95％ confidence interval (CI) was used to assess the strength of the association.Results Twenty-five eligible studies including 4 048 cases and 12 827 controls were identified.Meta-analysis showed a significant effect in the co-dominant model (T vs.C:OR 0 =1.25; Z=3.82,P=0.000;TTvs.CC:OR1=1.55,Z=3.79,P=0.000 ; except CT vs.CC:OR2=1.13; Z=1.78,P=0.076).Stratified analysis denoted that there was significantly higher risk in depression onset before 60 years old than depression onset after 60 years old.In subgroup analysis stratified by regressive meta-analysis,MTHFR C677T variant was statistically significantly relevant to depression risk in Chinese populations (OR 0 =1.58,Z=3.95,P=0.000; OR 1 =2.55,Z=3.92,P=0.000; OR 2 =1.43,Z=2.72,P=0.007),but not in no-Chinese populations (OR 0 =1.12,Z=2.10,P=0.035 ; OR 1 =1.23,Z=2.23,P=0.025 ; OR 2 =1.06,Z=0.74,P=0.456).Evidentiary strength rated by GRADE showed that 1 study moderate,3 studies low and 7 stdudies very low.Conclusion Our study provides evidence for a hereditary association between the MTHFR C677T variant and depression susceptibility,whereas there is an obvious variety in different regions,ethnicities and onset age.