To date, inter- and intra-laboratory consistency of binding assays for measuring anti-interferon (IFN)beta antibodies has not been assessed. In this investigation, two independent laboratories tested a library of 80 serum specimens obtained from multiple sclerosis (MS) patients treated with IFNbeta. For binding antibodies (BAbs) evaluations, each laboratory used both a capture-ELISA (cELISA) and an enzyme-immuno-assay (EIA), which is commercially available. Samples were also tested for neutralizing antibodies (NAbs). Data demonstrated good intra-laboratory reliability (r(pearson) > or = 0.86), and a good overall agreement between the results obtained from the two centers, using both the cELISA (69/80 of observed agreements) and the EIA (67/80). Accordingly, kappa coefficients (K) showed good concurrence (K > or = 0.651). There was also substantial agreement between cELISA and EIA measurements, as performed in both centers (Orbassano, 66/80, K = 0.631; Basel, 70/80, K = 0.717). However, by comparing NAbs and BAbs titers obtained with both assays, we found that a high degree of BAb-negative samples were positive in NAb-assay. Thus, our study does not support the usefulness of ELISA-based BAb assays as a screening tool for NAbs. Otherwise, BAb-assays can be used as a confirmation test, indicating that the decrease of the biological effects is due to antibodies. In this context, both ELISA-based assays are equally reliable techniques.