Although accumulating evidence strongly suggests that aplastic anemia (AA) is a T cell-mediated autoimmune disease, no target antigens have yet been described for AA. In autoimmune diseases, target autoantigens frequently induce not only cellular T-cell responses but also humoral B-cell responses. We hypothesized that the presence of antigen-specific autoantibodies could be used as a "surrogate marker" for the identification of target T-cell autoantigens in AA patients. We screened a human fetal liver library for serologic reactivity against hematopoietic stem/progenitor cell antigens and isolated 32 genes. In 7 of 18 AA patients, an immunoglobulin G (IgG) antibody response was detected to one of the genes, kinectin, which is expressed in all hematopoietic cell lineages tested including CD34+ cells. No response to kinectin was detected in healthy volunteers, multiply transfused non-AA patients, or patients with other autoimmune diseases. Epitope mapping of IgG autoantibodies against kinectin revealed that the responses to several of the epitopes were shared by different AA patients. Moreover, CD8+ cytotoxic T cells raised against kinectin-derived peptides suppressed the colony formation of granulocyte macrophage colony-forming units (CFU-GMs) in an HLA class I-restricted fashion. These results suggest that kinectin may be a candidate autoantigen that is involved in the pathophysiology of AA.