Antiphospholipid antibodies have been found to be strongly associated with syndromes characterised by spontaneous arterial and venous thromboses, recurrent miscarriage, immune thrombocytopenia, and occasionally neurological manifestations. These antibodies can be detected using solid phase immunoassays, and by their effect on prolonging phospholipid dependent clotting tests. This latter phenomenon is termed the lupus anticoagulant (LA). The relationship between anticardiolipin antibodies (ACA) and the LA activity of plasma was investigated in 14 patients. Plasma of these patients exhibited both LA activity and high levels of ACA. The patients included 7 with systemic lupus erythematosus, 6 without and 1 chlorpromazine induced lupus anticoagulant. 7 patients had a history of thrombosis and 7 did not, despite high antibody levels. Plasma was incubated in a serial fashion with solid phase cardiolipin and the residual ACA level and LA activity were monitored using a solid phase enzyme linked immunoassay, and the kaolin clotting time (KCT) and activated partial thromboplastin time (APTT) respectively. There was no correlation between baseline ACA levels and parameters of LA activity (dKCT or dAPTT) in contrast to previous reports. However, there was a concurrent reduction in both LA and ACA levels over 24 hours during incubation with cardiolipin in all patients. The rate of reduction of both parameters was highly correlated (r = 0.99. p less than 0.001). The relative reduction of LA activity versus ACA level varied between patients, and may represent different affinities for phospholipid in thromboplastin versus phospholipid in solid phase. Thus, despite the lack of concordance between LA and ACA in many patients, the two activities can be removed concurrently in vitro, suggesting similar binding specificities of the antibodies. The incomplete concordance could be explained by varying affinities for different structural presentations of the lipid antigen.