Although endothelial cell injury and microcirculatory intravascular thrombosis have been implicated in the pathophysiology of skin-flap failure, the basic underlying pathophysiology has not been documented previously. This study focuses on the morphologic changes and the alteration in platelet, fibrinogen, antithrombin III, and von Willebrand factor levels in flaps injured by arterial ischemia and reperfusion. A thrombogenic arterial anastomosis model is compared with simple arterial clamping as methods to achieve flap ischemia. Bilateral buttock skin flaps and latissimus dorsi island flaps were elevated in 12 pigs. All flaps had a primary ischemic insult of 2 hours' duration by simple clamp application. During this interval, a thrombus-generating, microvascular anastomosis was prepared, and during a 2-hour period of reperfusion, laser Doppler and transiluminator monitoring of the vascular pedicle allowed documentation of embolic events from the thrombus-generating anastomosis. In group 1 (n = 6), the flaps were then subjected to 7 (buttock skin) and 5 (latissimus dorsi) hours of complete arterial ischemia by clamping. During the secondary ischemic period, the poor microanastomosis was resected and repaired. Radioactively labeled autologous platelets (111In) and human fibrinogen (125I) were injected intravenously half an hour before secondary reperfusion. After 4 hours of reperfusion, flap biopsies and venous effluent were collected and prepared for electron microscopic analysis. The flaps and control tissue were harvested and the radioactivity was counted. In group 2 (n = 6), flaps were subjected to 6 hours of secondary ischemia by using the same technique as in group 1. Central venous and flap venous blood was sampled at baseline as well as upon immediate secondary reperfusion and after 4 and 8 hours of reperfusion. The hematocrit, platelet count, fibrinogen, antithrombin III, and von Willebrand factor levels were determined for these intervals. Platelets and fibrinogen accumulated significantly in buttock skin flaps and in the latissimus dorsi skin and muscle components as compared with similar control tissue (p < 0.05). There was no significant difference in platelet or fibrinogen accumulation after comparing the two ischemic models. Electron microscopic studies showed occluded capillaries with activated platelets in the flaps. Control tissue showed very little capillary occlusion. Platelet count was significantly decreased both in central venous (p < 0.05) and in adventitial infolding flap venous blood (p < 0.025) during immediate reperfusion as compared with baseline. These findings confirm that microcirculatory intravascular thrombosis is implicated in skin-flap ischemia-reperfusion injury. This study provides physiologic support for treatment modalities aimed at counteracting the various components in the coagulation pathways responsible for thrombus formation.