Forty-eight pediatric cadaveric renal transplantations, performed between May 1986 and February 1997, were retrospectively screened, pre- and post-transplant, for antibodies to human leukocyte antigen (anti-HLA) using complement-dependent cytotoxicity (CDC) assay and enzyme immunoassay (EIA). The correlation between anti-HLA immunization and graft outcome was investigated. The combined analysis of CDC and EIA enabled the differentiation between complement-fixing and non-complement-fixing, anti-HLA class I and anti-HLA class II antibodies. The median post-transplant follow-up for all patients with a functioning graft was 86 months (range 10-138 months). In the whole population, 16 grafts were lost: six following a non-immunologic complication; and 10 as a result of rejection. Of these 10 grafts lost, eight were in patients with pre- and/or post-transplant donor antigen specific (DAS) anti-HLA class I or class I + II antibodies; and two were in patients with DAS anti-HLA class II antibodies only. Three of these grafts were lost in patients with weak pre-existing DAS anti-HLA class I antibodies. Immunological graft loss appeared at a median post-transplant time of 38 months (range 2-68 months). All patients without DAS anti-HLA antibodies had a good graft outcome. The presence of pre- and post-transplant DAS anti-HLA antibodies, especially if directed against HLA class I, were associated with a poor graft outcome. A systematic search for, and identification of, anti-HLA antibodies should therefore be part of a pretransplant evaluation to allow the identification of 'unacceptable' donor HLA antigens, following which the impact of the HLA-cross-match on graft outcome will improve. Screening for DAS anti-HLA antibodies post-transplant could be helpful for detecting patients with an increased risk for graft loss following rejection episodes.