Background and objectives Anti-citrullinated peptide/protein antibodies (ACPA) have diagnostic significance and a pathological role in rheumatoid arthritis (RA). Multiple specificities of ACPA have been described, targeting various proteins such as filaggrin, fibrin, vimentin, collagen, enolase and some citrullinated peptides of Epstein-Barr virus nuclear antigen (EBNA) as well. Specificities as well as affinities of ACPA may influence its ability to trigger inflammatory cytokine production, modifying local and systemic inflammation in patients. Therefore we compared binding affinities of serum samples from 60 RA patients to citrullinated 19mer filaggrin peptide. Materials and methods The peptide specific IgG was affinity purified from sera of RA patients on insolubilized citrullinated peptides derived from filaggrin, fibrin, collagen, and vimentin. Peptide specific IgG concentrations in sera were calculated by ELISA, using a calibration curve prepared from the affinity purified IgG. Binding affinities of IgG and serum samples were tested on citrullinated peptide covalently coupled to the surface of ProteOn XPR36 GLH sensor chip. The 1:1 binding model was employed to calculate the apparent affinity constant, KD. Results 26 of the 60 serum samples have shown measurable binding affinity to citrullinated filaggrin peptide with KD values between 10–5–10–7M. In general, individual sera showing multiple specificities to citrullinated peptides in ELISA possess a higher KD to filaggrin peptide. An inverse correlation was observed between disease severity (DAS28 index) and binding affinities. Surprisingly, high degree of cross-reactivity was detected between affinity purified IgGs specific for citrullinated filaggrin, fibrin, vimentin and collagen. Conclusions Biosensor analysis has shown that ACPA are very heterogeneous regarding both specificities and affinities, furthermore, the affinity purified IgGs from individual ACPA are highly cross-reactive. Support: National Research, Development and Innovation Office (OTKA104846) and ELTE KMOP-4.2.1/B-10–2011–0002