A B S T R A C T A quantitative primate model of arterial thromboembolism has been characterized with respect to mechanism and usefulness in evaluating modifying variables. The model involved the kinetic measurements of 51Cr-platelets and 125I-fibrinogen consumption by femoral arteriovenous cannulae in chaired baboons.Cannula platelet consumption correlated directly with exposed cannular area for irradiated Silastic and polyurethane (correlation coefficients of 0.940 and 0.901, respectively; P <0.001) and remained steady state for months. Nonirradiated Silastic was only minimally reactive with platelets. Despite increased rates ofplatelet consumption circulating fibrinogen was not measurably destroyed by any of the cannulae tested. Cannula platelet consumption was independent of cannula flow rate, platelet count, heparin anticoagulation, and ancrod defibrinogenation."'In-platelet imaging of irradiated Silastic cannulae demonstrated luminal accumulation and subsequent embolization of irregular platelet masses. When irradiated Silastic cannulae were inserted as extension segments in the renal arteries of four animals the glomerular vessels became progressively occluded with nonfibrin-containing platelet thromboemboli. Nonirradiated Silastic cannulae in control arteries produced no significant vascular occlusion. Because the survival of platelets from animals with consumptive cannulae was not shortened in normal recipient animals we concluded that platelets were either irreversibly removed through thromboembolic consumption or unaffected in their viability.Oral administration of dipyridamole and sulfinpyrazone decreased cannula platelet consumption in a dosedependent manner with complete interruption at 20 and 250 ,umol/kg body wt per d (in three divided doses), respectively, whereas oral acetylsalicylic acid Received for publication 15 January 1979 and in revised form 4 April 1979. (10-330 ,umol/kg per d) had no measurable effect on cannula platelet consumption.We conclude that this primate model simulates arterial thrombotic processes in man and that this model is suitable for the in vivo evaluation of biomaterials and of drugs that modify platelet behavior.