Results from studies on the association of PTPN22 C1858T polymorphism with AITD risk are conflicting, we thereby perform this meta-analysis to derive a more precise effect on this possible association. Two investigators independently searched the PubMed, Embase, Wanfang and CNKI (China National Knowledge Infrastructure) databases. A total of 11 studies with 3764 AITDs cases and 3328 controls were finally identified. Statistically significant association was observed between PTPN22 C1858T polymorphism and AITD risk based on all studies (TT vs. CC, OR=2.18, 95%CI=1.31˜3.62; TC vs. CC, OR=1.50, 95%CI=1.29˜1.73; TT/TC vs. CC, OR=1.41, 95%CI=1.12˜1.78; TT vs. TC/CC, OR=2.00, 95%CI=1.21˜3.33). The results of subgroup analysis showed that: (1) regarding ethnic diversity, the variant genotypes TT/TC of C1858T were associated with a significantly increased AITD risk in Caucasians (TT/TC vs. CC, OR=1.41, 95%CI=1.09˜1.83) (2) regarding different countries, the statistically significantly association was observed in UK (TC vs. CC, OR=1.64, 95%CI=1.36˜1.98; TT/TC vs. CC, OR=1.65, 95%CI=1.37˜1.98) and other countries (including South Tunisia, Russia, Polish, Japanese) (TT vs. CC, OR=3.65, 95%CI=1.43˜9.33; TT vs. TC/CC, OR=3.41, 95%CI=1.34˜8.65). (3) regarding the subtypes of AITDs, patients with Graves' disease (GD) had a significant higher degree of C1858T polymorphism (TT vs. CC, OR=2.35, 95%CI=1.36˜4.05; TC vs. CC, OR=1.46, 95%CI=1.12˜1.89; TT/TC vs. CC, OR=1.54, 95%CI=1.33˜1.80; TT vs. TC/CC, OR=2.16, 95%CI=1.25˜3.72), while no association was observed in patients with Hashimoto's thyroiditis (HT). No publication bias was observed. Our results demonstrated that PTPN22 C1858T polymorphism was associated with AITD risk, especially in Caucasians.