Background Recurrent pregnancy loss (RPL) had said to be related to the angiotensin converting enzyme insertion/deletion polymorphisms (ACE I/D) gene polymorphisms. But the conclusions were controversial. This meta-analysis was conducted to investigate the real association in ACE I/D polymorphisms and RPL firstly. Methods Combine Pubmed Embase and HuGENet database in data analysis for this meta-analysis from October 2000 to November 2011. The metagen system was used to select the models and effects. Odds ratio (OR) with 95% confidence interval (CI) was used to assess the strength of this association. Results 9 studies from six countries with 1264 RPL and 845 controls were included according to our criterion. Following the metagen system, we used the dominant model with random effects. The summary OR 01.61 (95% CI: 1.10-2.36, I 2 0 59.0%), which suggested the ACE D allele might increase the RPL risk in Asia (OR01.97, 95% CI: 1.31-2.98, I 2 0 44.4%), among Asians (OR01.69, 95% CI: 1.06-2.36, I 2 032.7%). In additional, after conducting sensitivity analysis, the results had no differences except for Caucasian subgroup reached to the significance (OR02.059, 95% CI: 1.455-2.914), so we couldn't ignore the relationship between the polymorphisms of ACE D/I gene and Caucasians yet. There seemed no publication bias in our eligible studies with Begg's test (P 0 0.867). Conclusions Results in this meta-analysis presented the positive function of the ACE I/D polymorphism in increasing the RPL risk. Furfure prospective studies were needed to confirm the precise relationship between the ACE I/D and RPL.