The autoimmune hemolytic anemia of NZB mice is pathogenetically mediated by a genetically prescribed anti-erythrocyte autoantibody response directed to the X erythrocyte autoantigen. The cellular locus of the immunoregulatory defect underlying the anti-X response was explored by adoptively transferring bone marrow cells (BMC) from NZB mice to lethally irradiated histocompatible recipients. Before adoptive transfer, BMC from donor mice were assayed for antigen-binding lymphocytes with receptors for the X autoantigen (X-ABL) by immunocytoadherence assays and for anti-X autoantibody-secreting cells (X-PFC) by plaque-forming cell assays. Twelve weeks after adoptive transfer, splenic lymphocytes from recipient mice were assayed for X-PFC and humoral anti-X autoantibody by Coombs' tests. Transfer of 15 to 30 x 10(6) BMC containing 6 to 12 x 10(3) X-ABL but no X-PFC from 6- to 8-week-old NZB mice to lethally irradiated BALB/c, B10.D2, C57BL/Ks, and DBA/2 mice produced X-PFC in 70% of the recipients. Development of X-PFC was not simply dependent upon available X-ABL since transfer of 15-30 x 10(6) BMC, containing comparable numbers of X-ABL, from BALB/c, B10.D2, C57BL/Ks, or DBA/2 mice to NZB or syngeneic recipients did not produce X-PFC. Transfer of BMC from NZB mice to BALB/c, B10.D2, and DBA/2 mice with weekly administrations of AKR anti-theta antiserum had no effect on the development of X-PFC; Tlymphocyte ablation was evidenced by the absence of theta+ spleen cells. These results suggest that the pathogenetic anti-X response is not genetically prescribed at the level of macrophages, humoral factors, or T cells, but rather appears to be a phenotypic expression of a primary B lymphocyte defect permitting or promoting differentiation of NZB X-ABL.