The actual effect of HLA-DR matching in renal transplantation remains controversial. Since DNA analysis has been shown to be more reliable than serological typing, a re-evaluation of the impact of DR-matching on graft prognosis is required. In this study, 224 cadaver kidney transplantations performed in our center were retrospectively matched according to Restriction Fragment Length Polymorphism DR incompatibilities and compared to prospective serological DR-matching. Transplant outcome was evaluated using graft survival, first rejection onset and rejection frequency. In 18.8% individuals, a discrepancy between serology and DNA typing for at least one antigen was noted. Serology particularly failed to type recipients (21.7%) and 43.2% of the total missed antigens were serologically "blank" or unidentified ("X") alleles. A graft survival rate of 100% after one year was observed for transplantations with no DNA DR mismatch (n = 31). Furthermore, there was a definite correlation between DNA matching and (i), the percentage of individuals with one or more than one acute rejection episode (18% and 41.8% at one year for O incompatibility and pooled 1 and 2 incompatibilities respectively, p < 0.05); (ii), the mean of acute rejection per individual (p < 0.001); and (iii), the rejection onset time (p < 0.01). No correlation between serological matching and the acute rejection episodes parameters was noted. Since HLA typing could be performed in less than 2 hrs using new molecular biology techniques, we conclude that prospective DNA typing should improve kidney transplantation outcome in the near future.