Immune thrombocytopenia is frequently encountered in medical practice and is generally accepted as being caused by an IgG antibody. The capability of detecting platelet-bound IgG as a diagnostic and therapeutic modality is critical for appropriate care and management of patients with idiopathic thrombocytopenic purpura (ITP), as well as other immune thrombocytopenias. We have modified our previous assay (Br J Haematol 37:265, 1977) by employing protein A and PAP as a labeled antibody. Surface bound platelet IgG was quantitated by phase contrast microscopy after incubation with PAP, graded per 100 platelets and expressed as a reactive index (RI). Controls (n=13) had RIs ranging from 0.49 to 0.72 (mean 0.63 +/- 0.02 SE). The nonimmune thrombocytopenic group (n=7) had an RI ranging from 0.58 to 0.72 (mean 0.64 +/- 0.01 SE). In contrast, the immune thrombocytopenic group (n=28) had RIs ranging from 1.04 to 1.75 (mean 1.43 +/- ;0.03 SE). Platelet-associated IgG was evaluated further by absorbing representative sera samples from each group against washed granulocytes, red cells and platelets. Only when sera from the immune thrombocytopenic group were absorbed against platelets did the reactive indices of pre- and postabsorption samples change significantly. These findings suggest that our assay is clinically applicable in detecting platelet-associated IgG in immune thrombocytopenia and has the advantage of being rapid, reproducible and easy to perform in a clinical laboratory.