High positive anticardiolipin antibody tests have been associated with recurrent thrombosis and pregnancy loss. Although these antibodies were believed to bind negatively charged phospholipids, recent reports have suggested that a serum protein, beta 2-glycoprotein I (beta 2-GPI), may be the true antigen for these antibodies. To resolve this issue, we compared binding of 75 anticardiolipin-positive and 71 anticardiolipin-negative serum samples from patients with rheumatic diseases to beta 2-GPI by using an enzyme-linked immunosorbent assay (ELISA). Serum samples from 30 healthy blood donors and 10 laboratory personnel were used as normal controls. We found no difference in binding between the three groups of serum samples. In addition, when binding to beta 2-GPI coated plates was compared with binding to ELISA plates without beta 2-GPI (blank), no difference was observed. Finally, binding of anticardiolipin-positive serum samples to plates coated with cardiolipin-beta 2-GPI mixture varied directly with the cardiolipin concentrations. Based on these findings, we conclude that anticardiolipin-positive serum samples do not bind beta 2-GPI.