Antiphospholipid antibodies (aPL) are heterogeneous and are now accepted to be mainly phospholipid-protein-dependent antibodies. Although these antibodies are classically associated with thrombosis, their clinical relevance remains to be established. The subgroups of antibodies characterized by their proteic targets were reported to be more appropriate thrombotic markers. We analysed the prevalence of a large panel of antiphospholipid-related antibodies (aPLR), comprising antibodies directed to phospholipid-protein complexes and to different protein cofactors (beta2GPI, prothrombin, annexin V and protein S), in 122 consecutive unselected patients who had experienced at least one venous thrombotic event. The presence of lupus anticoagulants was assessed with an integrated assay using hexagonal phase phospholipids. Two types of aPL (APA and anti-beta2GPI-PL) were measured using a mixture of phospholipids containing cardiolipin and goat serum or human beta2GPI, respectively, as a source of protein cofactor. Our results show a similar prevalence, close to 15%, of lupus anticoagulants, APA and anti-beta2GPI-PL. In contrast, antibodies to beta2GPI were detected in only 8% of the patients, and very few patients had antibodies directed to other proteins. Of the 35 patients having at least one positive aPLR, 17 were classified as severe, because they had recurrent or early onset of thrombosis (< 35 years). The distribution of aPLR between severe and mild cases was not significantly different except for lupus anticoagulants. Our results clearly indicate that lupus anticoagulant is the only aPLR test to be strongly associated with the severity of thrombosis.