ABSTRACT. Evidence is accumulating that chronic inflammation has an important role in prostate cancer. Two common polymorphisms in the prostaglandin-endoperoxide synthase 2 (PTGS2) gene, rs20417 and rs689470, have been found to alter the risk for prostate cancer, but the various studies are not in agreement. To derive a more precise estimation of this association, all available studies were considered in a meta-analysis, with 10,700 patients and 13,021 controls for rs20417 and 4087 patients and 3761 controls for rs689470. We used odds ratios (ORs) to assess the strength of the association, and 95% confidence intervals (CIs) to determine the precision of the estimate. When all groups were pooled, we did not detect a PTGS2 polymorphisms and prostate cancer significant association of rs20417 polymorphism with prostate cancer risk. Similarly, no associations were found in the subgroup analysis. However, we found that rs689470 was significantly associated with a trend towards increased prostate cancer risk when using both additive (OR = 2.15, 95%CI = 1.04-4.44, P = 0.04) and recessive models (OR = 2.07, 95%CI = 1.07-4.03, P = 0.03) to analyze the data. In subgroup analyses stratified by ethnicity, there was no evidence that rs689470 has a significant association with prostate cancer in Caucasians. Based on our meta-analysis, rs689470 polymorphism is significantly associated with prostate cancer risk in the overall population. Nevertheless, we suggest that further studies should be made to confirm these findings.