We report a patient with an erythrocyte autoantibody in whose serum a broadly reactive antibody was transiently replaced by a monospecific autoanti-Jka. On preoperation evaluation, a 49-year-old man, who had never been transfused, exhibited both a positive antibody screen and a positive direct antiglobulin test. Although a broadly reactive antibody had been present 2 years earlier, only anti-Jka was found in the serum on preoperative testing. In contrast, an acid eluate prepared from the patient’s red cells at that time was reactive with all cells from a 10-cell panel – a finding consistent with the broadly reactive autoantibody noted earlier. Repeat testing of a sample obtained 1 week later revealed only a broadly reactive autoantibody in both serum and eluate. This patient is notable in two respects. He exhibited a rare autoantibody, monospecific anti-Jka, and the specificity of the autoantibody changed from broadly reactive when first detected to anti-Jka and then back to broadly reactive. Thus, antibodies directed against specific antigens, in the setting of autoimmune hemolytic anemia, may be part of the autoimmune process and not always alloantibodies.