Heparin-induced thrombosis is due to an immune-mediated activation of circulating platelets and has significant clinical implications for patients with vascular disease. The purpose of this article was (1) to define the biochemical mechanisms of heparin-induced platelet activation (HIPA) and (2) to determine the relationship between thromboxane A2 (TxA2) synthesis and platelet granule release. In two patients with confirmed HIPA, heparin (3 U/ml) induced extensive platelet aggregation (61.5%), release of 14C-serotonin (81.5% of releasable 14C-serotonin, a dense granule marker) and platelet factor 4 (63.7% of releasable platelet factor 4, an alpha granule marker) and generation of TxB2, a stable metabolite of TxA2 (100% relative to serum control). In one patient heparin did not induce release of n-acetyl-beta-glucosaminadase (N-AC, a lysosomal granule marker), and aspirin (4 mmol/L), which abolished TxA2 synthesis, prevented aggregation and granule release. In the second patient heparin did induce release of N-AC (39.7% of releasable N-AC) and aspirin, despite abolishing TxA2 synthesis, did not prevent aggregation or granule release. In contrast, by elevating intracellular cyclic adenosine monophosphate, iloprost (0.01 mumol/L), a stable prostacyclin analogue, prevented heparin-induced aggregation, granule release, and TxB2 generation in both patients. Thus we show (1) HIPA can proceed independently of TxA2 synthesis; (2) heparin in certain patients can release lysosomal hydrolases, thus mimicking strong platelet agonists such as thrombin; and (3) iloprost but not aspirin prevents HIPA regardless of the biochemical pathways involved.