Several variables that may affect accurate measurement of platelet-associated IgG (PA-IgG) were studied using a radioimmunoassay of the consumption type. The amount of PA-IgG of washed, unfixed normal donor platelets was 1.0 +/- 0.9 fg IgG/platelet (mean +/- 2 SD). Upon storage of washed platelets in a buffer containing EDTA, the amount decreased significantly to 0.2 +/- 0.2 fg IgG/platelet. Simultaneously, an increase in modal platelet volume was observed. Similar results were obtained when platelets were fixed with paraformaldehyde (PFA). We postulate that this decrease in PA-IgG is caused by the release of plasma IgG entrapped by the surface-connected canicular system of the platelet, when the platelets swell during storage in EDTA or fixation with PFA. This presence of varying amounts of entrapped plasma IgG may cause the wide discrepancies in PA-IgG found in normal donor platelets as well as platelets from ITP patients by other investigators. A good quantification of platelet-bound alloantibodies was possible with our assay when platelets were routinely fixed to diminish the amount of nonspecific PA-IgG. This was demonstrated with different anti-Zwa (= anti-PlA1), anti-Baka and anti-HLA sera. We also observed that fragments of platelets as well as fragments of cells of other types can cause aspecifically increased Pa-IgG values and can thus interfere with the proper measurement of platelet-bound antibodies in all kinds of immunoassays in general.