Glutathione peroxidase (GPX) is an endogenous antioxidant enzyme counteracting oxidative stress. Accumulating evidence has demonstrated that the GPX1 rs1050450 C > T polymorphism may modulate cancer risk, but the association of GPX1 rs1050450 polymorphism with bladder cancer (BC) and prostate cancer (PCa) is still inconclusive. This meta-analysis was designed to determine the exact association of GPX1 rs1050450 C > T polymorphism with the risk of bladder cancer and prostate cancer. Odds ratios (ORs) and 95% confidence intervals (CI) were calculated to estimate the association strength. Databases of PubMed, EMBASE, and China National Knowledge Infrastructure were searched to retrieve eligible studies. In total, ten eligible studies with 6,194 participants were included. By pooling all eligible studies, we found that carriers of the variant T allele were associated with a significantly increased risk of urinary tract cancer (T vs. C: OR = 1.459 and 95% CI, 1.086-1.962; CT/TT vs. CC: OR = 1.411 and 95 % CI, 1.053-1.891). In stratified analysis, we observed that the rs1050450 C > T polymorphism was significantly associated with an increased risk of BC (T vs. C: OR = 2.111 and 95% CI, 1.020-4.368; CT/TT vs. CC: OR = 1.876 and 95% CI, 1.011-3.480), while the association was not significant for PCa. Egger's test and Begg's test revealed no publication bias. The present meta-analysis provides evidence that the GPX1 rs1050450 C > T polymorphism leads to an increased risk of BC but not the risk of PCa.