Anti-phospholipid antibodies (aPL) are associated with an increased risk of thrombosis and recurrent fetal loss. Antibodies to prothrombin (aPT) have been associated with the anti-phospholipid syndrome (aPS). We assessed variations in aPT assay methodology to optimize an aPT method that was used to screen patients with aPS (n = 66). Detection of aPT using enzyme-linked immunosorbent assay was influenced by the concentration of the capture antigen, the microtitre plate type and the buffer system. The combination of gamma-irradiated plates, a phosphate-buffered saline buffer and coating antigen of 10 microg/ml prothrombin was the most sensitive. Both serum and citrate samples are suitable for the detection of aPT. Under these conditions aPT IgM but not IgG were found to be associated with thrombosis and/or fetal loss.