Forty-three biopsies were performed between 30 and 60 min after reperfusion. Patients (22 males/21 females, mean age 41 ±12 years, mean donor age 32 ±14 years) were treated either with antilymphocytic globulin, cyclosporin, and prednisolone (24 cases), or OKT3, cyclosporin, and prednisolone (19 cases). Ten patients had delayed post-transplant renal function (DPRF), defined as haemodialysis requirements after surgery, and seven patients had acute rejection 11 ±16 days post-transplant. Kidneys were perfused with a hypertonic solution containing mannitol. All patients were followed up for at least 30 months. During follow-up, five patients lost their grafts due chronic rejection, two patients due to non-compliance and one due to recurrence of focal seg-mental glomerulosclerosis. One patient died from heart infarction. Biopsies were stained with H&E, Masson's trichrome, periodic acid-Schiff (PAS) and silver methenamine. Interstitial fibrosis, interstitial oedema, tubular vacuolization, and peritubular capillary oedema were measured using a semiquantitative scale. Five 400 × magnification micrographs of cortical inter-stitium from silver-methenamine-stained sections were used to measure percentage of interstitial surface with a morphometer. Interstitial surface was 18.7 ±6.2% (range 3.2–35.3%). A positive correlation was found between interstitial surface and donor age (r= 0.469, P=0.0015). No relationship was found between warm and cold ischaemia times and tubular vacuolization or peritubular capillary oedema. Patients with DPRF had a significantly increased interstitial surface (23 ±8%) when compared with patients without DPRF (17 ±5%), (P=0.014). There was a positive relationship between interstitial surface and number of days required to achieve a plasma creatinine of 300 μmol/1 after surgery, this fitted an exponential curve (r=0.578, P=0.0012). Patients who had an episode of acute rejection were not included in this calculation. A positive correlation was also found between interstitial surface and plasma creatinine at 12 months (r=0.692, P=0.0001), 18 months (r=0.713, P=0.0001), and 24 months (r=0.586, P=0.0023) after surgery. Patients who lost their grafts during follow-up were not included in this calculation. The relationship between interstitial surface and plasma creatinine 30 months after transplantation was not significant. There was no relation between tubular vacuolization or peritubular capillary oedema and number of days required to achieve a plasma creatinine of 300 μmol/1 or plasma creatinine 12, 18, 24, and 30 months after transplantation. We conclude that assessment of donor renal biopsies may help to predict post-transplant renal function. The increase of interstitial surface due to pre-existing fibrosis is associated with poor post-transplant graft performance.