Antibodies against phospholipid antigens (APA) have been demonstrated in idiopathic thrombocytopenic purpura (ITP), but their clinical and pathogenetic significance has remained elusive. In this study we analyzed the prevalence and clinical features of ITP patients with elevated APA. In addition, we prospectively evaluated APA levels after treatment with corticosteroids and compared them with platelet- associated immunoglobulin (PAIgG) titers. We studied 149 patients with newly diagnosed ITP. Of these, 78 had a platelet count less than 50 x 10(9)/L and received an initial treatment with oral prednisone (PDN). In 71 asymptomatic cases with platelet counts between 50 x 10(9)/L and 120 x 10(9)/L, no therapy was scheduled. However, in five of them, the platelet count fell below 50 x 10(9)/L after more than 12 months; these patients were treated with PDN. Tests for APA included the measurement of anticardiolipin antibodies (ACA) with a solid-phase immunoassay and the detection of the lupus-like anticoagulant (LA) activity with coagulation tests that included kaolin-clotting time, dilute Russel's Viper venom time, activated partial thromboplastin time (aPTT), and dilute aPTT. Controls consisted of 174 apparently healthy subjects. Either LA or elevated ACA was seen in 69 patients (46.3%) at diagnosis. LA and ACA were both elevated in 24 cases (16.1% of the overall patient population and 34.8% of patients with high APA concentrations). No correlation was found between LA ratio values and ACA-IgG or -IgM titers, or between ACA-IgG and ACA-IgM levels. The presence of these antibodies was not associated with sex, age, platelet count, or the severity of hemorrhages. PAIgG was detected in 106 of 127 cases (83%). Again, no relationship was observed with clinical parameters or with APA levels. However, all cases with elevated APA also had increased PAIgG. With regard to the clinical course, we were not able to detect any significant difference between patients with normal and elevated APA. An initial complete response to prednisone treatment was observed in 43 of 83 cases (51.8%), with 13 (15.7%) achieving a prolonged complete remission. APA levels were not significantly modified after PDN therapy and on relapse. We conclude that APA positivity is a common finding in patients with ITP and does not select a category with different clinical features. APA levels are not influenced by immunosuppressive therapy with steroids and are not related to the activity of the disease. Therefore, we do not support a role for APA in the pathogenesis of ITP.