Objective The anti-carbamylated protein (CarP) antibody is a novel biomarker that might help in the diagnosis of rheumatoid arthritis (RA). We aim to assess the diagnostic value of anti-CarP antibody for RA. Methods We systematically searched PubMed, Embase, the Cochrane Library, Web of Science, and Scopus for studies published by December 15, 2015. Studies in any language that evaluated the utility of the anti-CarP antibody in the diagnosis of RA in which healthy donors or patients without arthritis or arthralgia served as controls were included. Two investigators independently evaluated studies for inclusion, assessed study quality and abstracted data. A bivariate mixed-effects model was used to summarize the diagnostic indexes from 7 eligible studies. Results The pooled sensitivity, specificity, and positive and negative likelihood ratios for anti-CarP antibody were 42% (95% CI, 38% to 45%), 96% (95% CI, 95% to 97%), 10.2 (95% CI, 7.5 to 13.9), and 0.61 (95% CI, 0.57 to 0.65), respectively. The summary diagnostic odds ratio was 17 (95% CI, 12 to 24), and the area under summary receiver operator characteristic curve was 80% (95% CI, 77% to 84%). Conclusion Anti-CarP antibody has a moderate value in the diagnosis of RA with high specificity but relatively low sensitivity.