Previous studies have explored the relationship of PTPN22 and TLR9 polymorphisms with systemic lupus erythematosus (SLE). In consideration of the population stratification, conflicting results and updating data, we conducted a comprehensive meta-analysis, which consists of a total of 17 research articles (9120 cases and 11,724 controls) for PTPN22 and 20 articles (including up to 2808 cases and 3386 controls) for TLR9. Significant association was verified between PTPN22 rs2476601 and SLE in the overall population (OR = 1.511 per T allele, 95% CI 1.338-1.706, P = 2.931 x 10(-11)) and under dominant model of T allele (TT+CT vs. CC: OR = 1.531, 95% CI 1.346-1.742, P = 9.17 x 10(-11)). Analysis after stratification by ethnicity indicated that PTPN22 rs2476601 was related to SLE in Americans (OR = 2.566, 95% CI 1.796-3.665, P = 2.219 x 10(-7)), Europeans (OR = 1.399, 95% CI 1.261-1.552, P = 2.153 x 10(-10)), and Africans (OR = 4.14, 95% CI 1.753-9.775, P = 1.0 x 10(-3)). We did not observe any association between TLR9 polymorphisms (rs187084, rs352140, rs5743836 and rs352139) and SLE under any model, after excluding the data that were inconsistent with Hardy-Weinberg equilibrium (HWE). In summary, PTPN22 rs2476601 was significantly interrelated with SLE and contributed to susceptibility and development of SLE in Americans, Europeans and Africans in this analysis, while their relationship needs to be validated in Africans by future research.