The aim of this study was to try to better characterize, on the basis of maternal platelet count and antiplatelet antibodies, women with ITP or a history thereof at risk of being delivered of a child affected by neonatal thrombocytopenia. Results show that either clinical classification or maternal platelet count were not effective in predicting the occurrence of neonatal thrombocytopenia. Effects of prednisone on platelet-bound and unbound antibodies were studied prospectively in 12 non-pregnant women with ITP; no increase of these parameters was observed after treatment. Thus, the risk of corticosteroid therapy in the management of pregnant ITP patients remains hypothetical and unproven.