The effect of different methods of plasma preparation on the results of 1) a clotting assay for lupus anticoagulant (LA) detection (the dilute activated partial thromboplastin time, dAPTT), and of 2) an ELISA test for anticephalin antibody (aCEPHA) detection, was evaluated. It is well known that platelet disintegration resulting from freeze-thawing of plasma samples may release procoagulant phospholipid--"LA-bypassing" activity. Even with fresh plasma, the dAPTT of LA positive samples was sensitive to the presence of residual blood platelets. This effect was accentuated by freezing and thawing: with test plasma that had been prepared by a centrifugation force of 3,000 g or less for 15 min at 4 degrees C, freeze-thawing caused a significant shortening of the dAPTT. This phenomenon could not be demonstrated with filtered test plasma, which was platelet free. Surprisingly, ultracentrifugation also led to a substantial shortening of the dAPTT compared to filtered plasma, and should not be recommended as a method of plasma preparation for LA detection. The ELISA test was less sensitive to residual platelets than the dAPTT. Thus, plasma prepared by a centrifugation force of at least 1,500 g may be stored at -20 degrees C before performance of the ELISA test. For the dAPTT, filtering of test plasma and control plasma after centrifugation is recommended for maximum sensitivity, regardless of whether they are to be examined in the fresh state or after freezing and thawing.