OBJECTIVES: This study evaluated the diagnostic performance of anti-carbamylated protein (anti-CarP) and anti-citrullinated fibrinogen (ACF) antibodies in rheumatoid arthritis (RA). METHODS: We searched the Pubmed, Embase, and Cochrane library databases, and performed two meta-analyses on the diagnostic accuracy of anti-CarP and ACF antibodies in patients with RA. RESULTS: The meta-analysis included data from 12 studies. Of these, seven studies, which included 1,749 patients and 1,044 controls, examined anti-CarP antibody, and five studies, which included 733 patients and 1,178 controls, examined ACF antibody. The pooled sensitivities and specificities of anti-CarP antibody were 43.9% (95% confidence interval [CI], 41.6~46.3) and 94.5% (95% CI, 93.0~95.8), respectively, and those of ACF antibody were 68.3% (95% CI, 64.9~71.6) and 95.8% (95% CI, 94.5~96.9), respectively. The positive likelihood ratio (PLR) of anti-CarP antibody were 9.901 (95% CI, 5.005~19.58), negative likelihood ratio (NLR) was 0.597 (95% CI, 0.541~0.658), and diagnostic odds ratio (DOR) was 14.64 (95% CI, 8.004~34.45). For ACF antibody, PLR was 16.14 (95% CI, 10.23~25.42), NLR was 0.292 (95% CI, 0.192~0.444), and DOR was 58.61 (95% CI, 26.61~129.1). There were no significant difference in sensitivity, specificity, PLR, NLR, AUC, or Q* index between ACF and anti-cyclic citrullinated peptide (anti-CCP) in the diagnosis of RA. CONCLUSION: Our meta-analysis demonstrates that both anti-CarP and ACF antibodies are highly specific for diagnosing RA. However, while ACF and anti-CCP showed comparably high diagnostic accuracy, anti-CarP antibody showed low sensitivity in diagnosing RA.