Objective. We compared the Disease Activity Score 28 (DAS28) using C-reactive protein (DAS28-CRP) with DAS28 using erythrocyte sedimentation rate (DAS28-ESR) in assessing rheumatoid arthritis (RA) activity and determining European League Against Rheumatism (EULAR) response criteria. Methods. We searched the PubMed, EMBASE, and Cochrane databases and performed a meta-analysis to examine comparisons between DAS28-CRP and DAS28-ESR by RA activity and EULAR response criteria. Results. A total of ten studies were included in this meta-analysis. Significantly more patients were classified as having remission or low disease activity when using DAS28-CRP than when using DAS28-ESR (odds ratio [OR]=1.869, 95% confidence interval [CI]=1.180 to 2.959, p=0.008; OR=1.411, 95% CI=1.256 to 1.586, p=7.0×10 −8 ), whereas fewer patients were classified as having high disease activity when using DAS28-CRP than when using DAS28-ESR (OR=0.534, 95% CI=0.388 to 0.734, p=1.1×10 −4 ). More patients were classified as having good response with criteria were based on DAS28-CRP than with DAS28-ESR (OR=1.390, 95% CI=1.183 to 1.632, p=6.10×10 −5 ). Conclusion. Our meta-analysis demonstrates that DAS28-CRP underestimates disease activity and overestimates response by the EULAR response criteria compared to DAS28-ESR. (J Rheum Dis 2016;23:241-249)