Glioma is defined as a common brain tumor which causes severe disability or death. As many genes are reported to relate with glioma’s occurrence and development, their prognostic and therapeutic value still remains uncertain. This study aimed at investigating the association between STAT3/p-STAT3 and glioma prognosis. Nine studies (12 trials) scored ≥5 on the Newcastle-Ottawa scale were meta-analysed from the Medline, Embase, and Web of Science databases. We found that STAT3/p-STAT3 overexpression in glioma patients was associated with worse overall survival ( hazard     ratio     HR = 1.40 , 95 % confidence     interval     CI = 1.05 ~ 1.86 , P = 0.020 ), progression-free survival ( HR = 2.05 , 95 % CI = 1.63 ~ 2.58 , P < 0.001 ), and better recurrence-free survival ( HR = 0.37 , 95 % CI = 0.15 ~ 0.95 , P < 0.039 ). Subgroup analysis implied that STAT3/p-STAT3 overexpression was associated with worse OS in standard treatment ( HR = 1.80 , 95 % CI = 1.06 ~ 3.04 , P = 0.030 ), and in China ( HR = 2.18 , 95 % CI = 1.77 ~ 2.70 , P < 0.001 ), and metaregression analysis indicated countries ( P = 0.001 ) may be the source of heterogeneity in our study. In conclusion, we suggested STAT3/p-STAT3 was associated with poor prognosis in patients with glioma, which indicated that STAT3/p-STAT3 might be a valuable prognostic biomarker and a promising therapeutic target for glioma.