Flow cytometry (FC) provides a reproducible investigation of cell surface antigens on platelets. The aim of this study was to elaborate appropriate protocols and to compare them with other techniques that have already been published. (1) Venipuncture with tubes containing citrate was better for the preservation of the antigenicity than using ACD tubes. The isolated platelets could not be completely distinguished from detritus and protein aggregates. Therefore a platelet concentration between 10(7) and 10(8)/ml measurement buffer was necessary to obtain a sufficient resolution by FC. (2) Isolation methods using either differential centrifugation or diluted Ficoll-Hypaque as a flotation medium provided platelets of equal purity. The method with Ficoll-Hypaque resulted in a higher number of isolated platelets than differential centrifugation. The demonstration of platelets and their antigens in whole blood without isolation gave good results provided the platelets were not activated. Activation of platelets with 1 NIH-U thrombin/l resulted in the loss of a part of the highly activated platelets because of their aggregation. (3) Comparing different concentrations of paraformaldehyde in PBS, fixation with 1% for 15 min provided the best antigen preservation for most of the antigens investigated. Isolation induced platelet activation. In order to avoid this effect, the whole anticoagulated blood was fixed with 1% paraformaldehyde for 15 min immediately after venipuncture. Then the platelets were isolated using diluted Ficoll-Hypaque. In this way, systemic activation of platelets can be detected with antibodies against glycoproteins which are translocated from the alpha-granules or lysosomes to the cell membrane. These activation markers can be determined on immediately fixed platelets (already in the whole blood) without any interference due to unspecific activation caused by the isolation procedure. (4) Platelet treatment with citric acid at pH 3, in order to remove the antigenicity of HLA-class I molecules, was sensitive to immediate fixation with paraformaldehyde in the whole blood. Fixation after isolating the platelets made it possible to demonstrate antigen stripping, and the free heavy chain, devoid of the beta 2-microglobulin, could be clearly demonstrated. (5) Using standardization beads, the average number of antigenic sites per platelet could be determined for the investigated specificities. It was shown that antibodies which have been directly conjugated or biotinylated and combined with streptavidin-phycoerythrin yielded similar results in terms of the number of antigenic binding sites while unconjugated antibodies in combination with FITC-conjugated anti-mouse-IgG led to overestimation of antigenic binding sites.