OBJECTIVE: To evaluate the performing technical and clinical characteristics of an automated system for routine measurement of anticyclic citrullinated peptide antibodies (aCCP), a new marker for rheumatoid arthritis (RA). MATERIAL AND METHODS: Reproducibility, repeatability and linearity of aCCP, as measured by an automated fluorescent enzyme immunoassay (FEIA/Phadia), were evaluated and compared with the performance of a manual ELISA technique (Axis Shield Diagnostics). Clinical verification of both methods included estimation of sensitivity in RA patients (n = 42) and specificity in well-characterized non-RA autoimmune disease controls (n = 49) and healthy subjects (n = 39). RESULTS: Precision studies showed a coefficient of variation between 4.9 % and 10 % for the FEIA technique and between 6.35% and 19% for the ELISA technique. Both systems showed good linear response. Sensitivity of aCCP for RA was 74% for FEIA and 79% for ELISA. Specificity was 100% for both methods, as calculated for healthy subjects. For non-RA-diseased controls, specificities of 98% and 94% were obtained for FEIA and ELISA, respectively. Both methods were concordant in 97% of cases. Increasing the cut-off for the ELISA system from >5 U/mL to >11 U/mL resulted in lower sensitivity (71.4%) but higher specificity (98.0%), i.e. improved discriminating power between RA and non-RA and 100% agreement between both methods. CONCLUSION: Automated FEIA measurement of aCCP in the routine clinical laboratory improves imprecision compared to the manual ELISA. Our preliminary results suggest that an increase in cut-off for the ELISA can improve specificity to RA from 94% to 98 %