Platelet integrin αIIbβ3 (GPIIb/IIIa) plays a central role in the initiation of arterial thrombosis, but its contribution to disseminated microvascular thrombosis is less well defined. Therefore, wild-type mice (β3^+/+), β3-integrin–deficient mice (β3^−/−), and wild-type mice treated with a hamster monoclonal antibody (1B5) that blocks murine αIIbβ3 function were tested in models of large-vessel and microvascular thrombosis. In the large-vessel model, ferric chloride was used to injure the carotid artery, and the time to thrombosis was measured. In β3^+/+mice, the median time to occlusion was 6.7 minutes, whereas occlusion did not occur in any of the β3^−/− mice tested (P < .001). Fab and F(ab')₂ fragments of 1B5 increased the median time to occlusion. To initiate systemic intravascular thrombosis, prothrombotic agents were administered intravenously, and platelet thrombus formation was monitored by the decrease in circulating platelet count. Three minutes after the injection of adenosine diphosphate (ADP), collagen + epinephrine, or tissue factor, the platelet counts in β3^+/+ mice decreased by 289, 424, and 429 × 10³/μL, respectively. β3^−/− mice and wild-type mice pretreated with 1B5 Fab (1 mg/kg, IP) were nearly completely protected from the effects of ADP. In contrast, β3^−/− mice were only partially protected from the effects of collagen + epinephrine and minimally protected from the effects of tissue factor. In all cases, less fibrin became deposited in the lungs of β3^−/− mice than in wild-type mice. These results suggest that though αIIbβ3 plays a dominant role in large-vessel thrombosis, it plays a variable role in systemic intravascular thrombosis.