The effect of the 675 insertion/deletion (4G/5G) polymorphism of plasminogen activator inhibitor-1 (PAI-1) gene on the risk of venous thromboembolism (VTE) remains controversial. In this study, we performed a meta-analysis of published data regarding this issue. A comprehensive electronic search was carried out up until September 2006. A total of 22 articles were included in the analysis that was performed using random effects models. Eighteen papers, concerning patients without another known risk factor, comprised 2,644 cases and 3,739 controls. The alleles contrast (4G vs. 5G allele) yielded a statistically significant odds ratio (OR) of 1.153 (95% confidence interval [CI]: 1.068-1.246). In a sub-analysis of five studies that included 256 cases with another genetic risk factor and 147 controls, the combined per-allele OR was still significant (OR: 1.833,95% CI: 1.325-2.536). On the contrary, the analysis of five studies regarding cases with a non-genetic risk factor for VTE (antiphospholipid antibody syndrome, Behcet disease) provided insignificant results in all aspects. There was no evidence for heterogeneity and publication bias in all analyses. Based on our findings, the 4G allele appears to increase the risk of venous thrombosis, particularly in subjects with other genetic thrombophilic defects. Recommendation for detection of this polymorphism in evaluating thrombophilia in such patients might be considered.