Increasing studies investigating the association between steroid 5-alpha reductase type II gene polymorphism at codon 89 (SRD5A2 V89L) and susceptibility to prostate cancer (PCa) confer inconsistent results. To precisely estimate the relationship with more statistical power, a meta-analysis was performed. A comprehensive search was conducted to identify all case-control studies investigating such an association. Odds ratio (OR) and its 95% confidence interval (CI) were used to evaluate the size effect. Twenty-five eligible reports were identified including 8,615 cases/9,089 controls in 33 comparisons. In overall analysis, no significant associations were found in all genetic models. Subgroup analyses by ethnicity revealed that small excess PCa risks were observed in dominant model (OR, 1.11; 95% CI, 1.03-1.19 for (LL + VL) vs. VV; P < 0.01; P(heterogeneity) = 0.49) and L allele frequency comparison (OR, 1.09; 1.03-1.15 for L allele frequency; P < 0.01; P(heterogeneity) = 0.07) in Europeans. Meanwhile, SRD5A2 V89L polymorphism was significantly associated with an increased PCa risk in men aged < or =65 under the co-dominant (OR, 1.70; 95% CI, 1.09-2.66 for LL vs. VV; P = 0.02; P(heterogeneity) = 0.31) and recessive (OR, 1.75; 95% CI, 1.14-2.68 for LL vs. (VV + VL); P = 0.01; P(heterogeneity) = 0.12) models. However, no significant associations were found in Asians and Africans. Our study suggests SRD5A2 V89L polymorphism could play a low-penetrant role in PCa risk among Europeans and individuals younger than 65 years. Additional well-designed studies are warranted to validate these findings.