We think that the determination of acceptable mismatches is an efficient approach to increase the chance of finding a crossmatch-negative donor for highly sensitized patients. This approach has several advantages (4,5) namely: 1. There is no need for distribution of patient sera to other tissue typing centers. 2. Rather than performing crossmatches with all donors, most of which will be positive, selection is based on a predictable negative crossmatch. 3. Selection of potential donors is based on data obtained by the recipient centers, which has all the information concerning the patient's immunological background (transfusion, specific alloantibodies, autoantibodies) rather than on a negative crossmatch at another tissue typing center. 4. Selection is based on HLA-DR compatibility between donor and recipient, which is, both in our and other analyses (6), important for an optimal prognosis of graft survival in highly immunized patients. Of course, there are still patients for whom this protocol is not very helpful (rare HLA-types, hardly any or no acceptable mismatches). For these patients it might be that other approaches, including removal of circulating antibodies (7), may be the only solution.