Historical HLA class II serological typing results of transplantations performed in "The Leuven Collaborative Group for Transplantation" were subjected to retrospective Restriction Fragment Length Polymorfism (RFLP) DNA control typing by the Collaborative Transplant Study (CTS) DNA project using Polymerase Chain Reaction (PCR)-based DNA methods. We re-evaluated the serology/ RFLP-discrepant CTS DNA data for our local patients transplanted during a historical period (January 1988 until May 1992) before any class II DNA typing was performed in our tissue typing laboratory. These retyping results confirm both the CTS data for patient typing and the Eurotransplant data for donor typing. A confirmed high discrepancy rate of 19.0% (after exclusion of 2.2% transcription errors) was found in the patient population. A low discrepancy rate of 6.8% (after exclusion of 2.2% transcription errors) for the donor population is concordant with the Eurotransplant donor data. Only 4 of the 588 individuals were found to be incorrectly typed by the RFLP method; all involving the specificities DRB1*1102. This indicates that RFLP typing, as performed by the CTS DNA project, can be considered a valid, retrospective DNA typing system for the accurate interpretation of class II matching in organ transplantation. A second conclusion to be drawn from this study is the need for prospective DNA typing for kidney transplant recipients, as the discrepancy rate in this cohort is high. Our results suggest that with good quality serological HLA-DR typing, prospective donor DNA typing is not urgently needed.