The supply of deceased donor kidneys is inadequate to meet demand. To expand the pool of potential donors, ABO-incompatible transplants from living donors have been performed. We present the Mayo Clinic experience with such transplants. Enrollment was open to patients when the only available potential living kidney donor was ABO-incompatible. Conditioning consisted of plasma exchanges followed by intravenous immunoglobulin. Splenectomy was performed at the time of transplant surgery. Post-transplant immunosuppression consisted of anti-T lymphocyte antibody, tacrolimus, mycophenolate mofetil, and prednisone. Isoagglutinin titers and scores were determined before and after each plasma exchange. Transplant outcomes were determined. Twenty-six ABO-incompatible transplants were performed. No hyperacute rejection occurred. Mean patient follow-up was 400 days. Patient and graft survivals at last follow-up were 92 and 85%, respectively. Antibody-mediated rejection occurred in 46% and was apparently reversed in 83% by plasma exchange and increased immunosuppression. The initial plasma exchange reduced immediate spin and AHG hemagglutination reactivity scores by 53.5 and 34.6%, respectively. Over the course of the pretransplant plasma exchanges, the immediate spin and AHG hemagglutination reactivity scores decreased by 96.4 and 68.5%, respectively. At 3 and 12 months, the immediate spin and AHG hemagglutinin reactivity scores and titers were less than those at baseline but greater than or equal to those on the day of transplantation. Despite an increase in scores and titers, antibody-mediated rejection was not present. Pre-transplant plasma exchange conditioning combined with other immunosuppressives can be used to prepare patients for ABO-incompatible kidney transplantation from living donors, but antibody-mediated rejection post-transplant is a common occurrence and allograft survival may be reduced. Controlled clinical trials are needed to identify the optimum conditioning for ABO-incompatible renal transplants.