Antiphospholipid antibodies (aPL) characterize patients at risk for both arterial and venous thrombotic complications. Recently it has been recognized that the presence of plasma proteins such as beta 2-glycoprotein I(beta 2 GPI) and prothrombin are essential for the binding of aPL to phospholipids and that these proteins are probably the real target of aPL. The discovery of these new antigens for aPL introduces the possibility of new assays to detect the presence of aPL. However, it is not known whether these assays improve the identification of patients at risk for thrombosis. In this retrospective study we compared the value of the classic assays LAC (lupus anticoagulant) and ACA (anticardiolipin antibodies) to detect aPL associated with thrombotic complications, with new assays which are based on the binding of aPL to the plasma proteins prothrombin and beta 2GPI. To do so, we have used these assays in a group of 175 SLE patients and correlated the positivity of the different assays with the presence of a history of venous and arterial thrombosis. Control groups were patients without SLE but with LAC and/or ACA and thrombosis (n = 23), patients with thrombosis without LAC and ACA (n = 40) and 42 healthy controls. In the univariate analysis, in which no distinction has been made between high and low antibody levels, we confirmed LAC and ACA to be related to both arterial and venous thrombosis. Anti-beta 2GPI- and anti-prothrombin-antibodies, both IgG and IgM correlate with venous thrombosis and anti-beta 2GPI-IgM with arterial thrombosis. Multivariate analysis showed that LAC is the strongest risk factor (OR 9.77; 95% CI 1.74-31.15) for arterial thrombosis. None of the other factors is a significant additional risk factor. For venous thrombosis LAC is the strongest risk factor (OR 6.55; 95% CI 2.36-18.17), but ACA-IgM above 20 MPL units also appeared to be a significant (p = 0.0159) risk factor (OR 3.90; 95% CI 1.29-11.80). Furthermore, the presence of anti-beta 2GPI- and/or anti-prothrombin-antibodies in LAC positive patients (n = 60) does not increase the risk for thrombosis. The results showed that (i) the LAC assay correlates best with a history of both arterial and venous thrombosis and (ii) neither the anti-beta 2GPI ELISA nor the anti-prothrombin ELISA gives additional information for a thrombotic risk in SLE patients.