The objective of the study was to determine if it is justified to use the scarce resources of cadaveric kidneys on HLA-sensitized patients, by reviewing the initial and long-term outcome of cadaveric renal transplantation at Uppsala University Hospital, Sweden. Between January 1988 and December 1994, 402 renal transplantations were performed. The patients were divided into one group of sensitized recipients (peak panel antibody reactivity > or = 25%; n = 84) and a second of non-sensitized recipients (panel reactive antibodies < 25%; n = 318). The groups were comparable in terms of recipient and donor age, gender, HLA-A, -B and -DR mismatches and numbers of diabetics. None of the sensitized patients received a six-antigen-matched kidney. For the non-sensitized group, life table analysis showed a 1-year actuarial graft survival (GS) of 91.8% and a 4-year GS of 84.4%. The corresponding GSs for the sensitized group were 79.9% and 68.7%, respectively (P < 0.01). The statistical significance vanished if patients with primary non-function were excluded. When excluding donors above 55 years of age, kidneys with cold ischemia time above 20 h, and two-antigen (HLA-DR) mismatches, there was no detectable difference between the non-sensitized and sensitized groups at 1-year or 4-year GS. Although there is a statistical significance in GS between non-sensitized and sensitized recipients of a kidney transplant, this does not differ from other risk groups such as diabetics, rheumatoid disease sufferers or elderly recipients. We therefore conclude that the sensitized patient should be accepted on the waiting list for a kidney transplant and that it is worthwhile to do the utmost to transplant this category of patients. Our data indicate that kidney GS in sensitized recipients is more affected by negative risk factors such as older donors, long cold ischemia time and two-antigen HLA-DR mismatch, than the non-sensitized recipient. To improve the outcome, those negative factors should be avoided or reduced.