The implantation and one hour post-transplant renal biopsies from three types of allograft recipients were compared with a blind grading system: (1) 25 cadaver kidneys preserved by pulsatile perfusion, (2) seven cadaver kidneys preserved by simple hypothermia following electrolyte solution flush, (3) 18 kidneys from living-related donors. Significant lesions were found only in cadaver kidneys which had received pulsatile preservation. Microscopic findings were correlated with perfusing agent, length of perfusion and its characteristics, and subsequent clinical course of the patient. Perfusion-related injury was found to be morphologically identical to hyperacute rejection, although the lesion is produced by quite different mechanisms. Pulsatile preservation appears to be associated with a spectrum of mechanical endothelial injury ranging from minute breaks visible only ultrastructurally to areas of complete denudation baring the basement membrane. The exposed collagen activates the clotting sequence resulting in platelet and fibrin deposition, whereas in classical hyperacute rejection the triggering mechanism is cytotoxic recipient antibody. The extent of perfusion-related injury correlates well with length of preservation, quantity of fibrin deposited, and, most importantly, with both the immediate and long-term post-transplant failure rate. In some patients the injury appears to be produced by cytotoxic antibodies in the plasma perfusate, which combine with antigens in the kidney ex vivo. The Ag-Ab complex activates complement and coagulation sequences in vivo after reimplantation. Early results with albumin or purified plasma fraction perfusates suggest this portion of perfusion-related injury can be eliminated. Comparison of pre- and postimplantation biopsies of the kidneys preserved by simple hypothermia or by pulsatile preservation suggests that perfusion-related injury is much more common than is true hyperacute rejection mediated by recipient cytotoxic antibodies. We suggest that the term "hyperacute rejection" be reserved for situations where significant endothelial drainage has been excluded by preimplantation biopsy and where recipient cytotoxic antibodies can be proved.