Hyperacute renal allograft rejection is initiated by primary immune injury to vascular endothelium and is propagated by secondary vasoconstriction, platelet aggregation and intravascular coagulation. Previous dissociation of these primary and secondary events, with graft survival in one human, suggested that the more usual graft failure was due to secondary injury. As a basis for further modification studies, this primate model most closely resembled its counterpart in man, as the onset and intensity of functional, morphologic and biochemical alterations were similar. Unmodified allografts failed within 5 minutes. The earliest and most abnormal finding was marked reduction in renal blood flow affecting all compartments. By 5 minutes, histologic changes of hyperacute rejection as well as antibody and faint C3 deposits were noted, but biopsies suggested that the initial flow reduction was more likely due to vasoconstriction, which was then followed by vascular obstruction. Glomeruli appeared most damaged, but at the highest antibody titer arterial injury was more prominent. Early red cell sequestration and stasis was marked, followed by progressive platelet clumping and neutrophil infiltration. While the decline in renal venous C3 levels was prompt, as in man, early intrarenal activation of the coagulation, fibrinolytic and kinin-forming systems could not be demonstrated, and fibrin formation was sparse by light and fluorescence microscopy. Qualitatively similar histologic and functional alterations were noted in autograft controls. While the initiating event was unclear and may have been accentuated by the arteriovenous shunts utilized, the final mechanism was probably marked vasoconstriction with renal ischemia. Intrarenal C3 consumption was an important finding and was not associated with tissue deposits of antibody or complement; it may provide a parallel with the progressive complement-mediated injury associated with acute myocardial ischemia noted by others. Endothelial injury was not seen in arteries, and all alterations were delayed in onset and progressed more slowly than in allografts. These findings may elucidate the mechanism of early malfunction of most autografts. Treatment of additional autografts with increasing doses of heparin progressively reversed these changes and even prevented the initial reduction in blood flow. Therefore, many alterations consistent with hyperacute rejection which are probably immediately responsible for graft failure can also be initiated by nonspecific, nonimmunologic events and, where injury is less intense, can be prevented pharmacologically. This model provides a means of dissecting the injurious events and subsequent evaluation of the effectiveness and interaction of various agents on the damaging secondary alterations which occur during hyperacute rejection.