The identification of the presence of antiphospholipid in plasma is recognised to be of diagnostic and prognostic importance in subjects with thrombotic disease, recurrent miscarriage or collagen vascular disorders. A number of coagulation assays are currently employed for the detection of lupus anticoagulant (LA), many of which are influenced by reagent dependent and methodological variables. In the present study lyophilised plasma samples from three subjects with "strong", "weak" and "absent" LA were tested in 220 centres. The most commonly used tests for LA were Activated Partial Thromboplastin Time (APTT), Dilute Russell Viper Venom Time (DRVVT) and Kaolin CLotting Time (KCT). Median DRVVT ratios were 1.75, 1.17 and 1.10 for the three samples. The presence of a strong LA was not detected by 4% of laboratories. The correct diagnosis was made by 94% of users of DRVVT and 85% of users of KCT. A weak LA was not detected by over half of centres. Correction was observed on addition of plasma and also in platelet neutralisation. The correct diagnosis was made by 37% of users of DRVVT and 27% of users of KCT. Lupus Anticoagulant was falsely considered to be present in a Factor IX deficient plasma by approximately one quarter of laboratories. Amongst users of DRVVT and KCT absence of LA in this sample was correctly reported by 73% and 69% of centres respectively. The accuracy of testing for LA in the present study is suboptimal and this is likely to have important clinical consequences. There is clearly a need for greater conformity in the selection and performance of LA tests to facilitate accurate diagnosis of this important group of disorders.