Previous surveys had evaluated the effects of the PD-1, CTLA-4 gene polymorphisms on susceptibility to ankylosing spondylitis (AS), but the results remained controversial. To briefly examine these consequences, a comprehensive meta-analysis was conducted to estimate the relationships between PD-1 rs11568821, rs2227982, rs2227981, CTLA-4 +49 A/G and -318 C/T polymorphisms and AS risk. The available articles dated to December 2014 were searched in the PUBMED, MEDLINE and EMBASE databases. The data of the genotypes and/or alleles for the PD-1 rs11568821, rs2227982, rs2227981, CTLA-4 +49 A/G and -318 C/T polymorphisms in the AS and control subjects were extracted, and statistical analysis was conducted by STATA 11.2 software. Summary odds ratios (ORs) with their 95 % confidence intervals (95 % CIs) were calculated to determine the strength of associations with fixed-effects or random-effects models. A total of eight published studies were finally involved in this meta-analysis. Meta-analysis of PD-1 rs2227982 polymorphism under the T allele versus C allele (OR 1.744, 95 % CI 1.477-2.059, P < 0.0001), TT+TC versus CC (OR 2.292, 95 % CI 1.654-3.175, P < 0.0001), TT versus CC (OR 1.883, 95 % CI 1.299-2.729, P = 0.001) revealed a significant association with AS. Our meta-analysis demonstrated that the rs2227982 polymorphism in the PD-1 gene might contribute to AS susceptibility. However, further studies with large sample sizes and among different ethnicity populations should be required to confirm this association.