The incidence of acute tubular necrosis ATN after cadaveric kidney transplantation in our centre has been in the range of 50%. A prospective study was carried out in 1991 and 1992 to assess the effect of in situ perfusion and hypothermic storage of kidneys harvested from brain-dead haemodynamically stable and unstable donors. Three litres of Ringer's solution were used for in situ perfusion. In 40 cases, the kidneys were stored in Euro-Collins (EC) solution and in the other 78 cases, in University of Wisconsin (UW) solution. Among the factors that could contribute to ATN, we analysed warm ischaemia time, anastomosis time and cold storage time. Function was considered to be delayed if the patient required posttransplantation dialysis. The donors were considered haemodynamically unstable when hypotension before harvesting was present (BP < 70 mm Hg over 2 h) despite high doses (> 15 microg/kg per minute) of dopamine or when cardiac arrest occurred at the time of harvesting and oliguria had been present for at least 2 h. Haemodynamically stable donors with a BP greater than 80 mm Hg had a normal diuresis. In all donors in this group the dose of dopamine was lower than 10 microg/kg per minute. The study showed that storage in UW solution did not influence the incidence of ATN in kidneys harvested from haemodynamically unstable donors. Differences observed in our study were due to haemodynamic status preceding donor nephrectomy and length of cold storage time.