We sought to verify earlier reports of increased platelet reactivity in patients with peripheral arterial disease (PAD) during perioperative heparin administration, and to test the hypothesis of platelet hypersensitivity to heparin in these patients. Before and after incubation of platelet rich plasma with unfractionated (UH), low molecular weight heparin (LMWH), and a low molecular weight heparinoid, real-time quantitative assessment of platelet function was performed by stagnation point flow adhesio-aggregometry (SPAA) in 21 patients with PAD and 14 healthy volunteers. With SPAA the occurrence of spontaneous aggregation is pathological. In the 15 patients requiring operation, platelet function and count were measured at regular intervals. To detect heparin dependent antibodies, the heparin induced platelet activation assay (HIPA) was performed preoperatively and after 10 days of heparin therapy. Mean baseline platelet adhesion in patients was double that observed in controls (p < 0.001). Spontaneous aggregation was seen in 9 (43%) patients and no controls (p < 0.001). In controls heparinoid reduced, whereas UH and LMWH slightly increased adhesion. Spontaneous aggregation was observed once with UH. Platelets from patients showed significantly enhanced adhesiveness and aggregability (p < 0.05) with UH and LMWH when compared to controls. Effects with the heparinoid were less pronounced and non-significant. In patients requiring operation, postoperative increases in platelet function and reductions in count were significant (p < 0.001). Ten (67%) experienced a fall in platelet count of > 50%. Preoperatively the HIPA assay showed no evidence of antibodies, whereas after heparin administration antibodies were verified in 4 (32%) patients and could not be ruled out in 6 (40%). Three developed postoperative thrombosis, in one case fatal. A hypersensitive in vitro and in vivo platelet response to heparin was verified in patients with PAD and a large number developed the immunological type of heparin-associated thrombocytopenia. Our findings suggest that a thrombin antagonist which does not interact with platelets may give the best perioperative protection in these patients.