The IA mutant mouse strain, B6.C-H-2bm12 (bm12) has been used to address several important questions for the role of Ia molecules in immune responses to foreign antigens. Numerous publications using bm12 mice have led to conclusions concerning (1) the number and relative importance of functional sites on Ia molecules; (2) the effects of qualitative versus quantitative differences in Ia; (3) whether T cells recognize Ia sequence or conformation; and (4) if gene conversion events, such as the one that putatively occurred in bm12, transfer functional Ir gene epitopes. Because of the importance of these conclusions, as well as their controversial nature, we have undertaken a comprehensive and systematic analysis of the aberrant immune response of bm12 mice to heterologous insulin. Responses to beef, horse, and sheep insulin were compared in B6 and bm12 mice by T cell proliferation, enumeration of plaque-forming cells, and quantitation of serum antibody levels. Various doses of antigen were administered and the kinetics of each response was monitored at various times. The findings of these studies suggest (1) B6 and bm12 mice both mount comparably high levels of response to sheep and horse insulins; (2) in contrast to the good response of B6 mice to beef insulin, bm12 mice showed dramatically impaired responses, as evident from both the lower magnitude of the response in all three assays as well as the difference in the kinetics of the response in B6 and bm12 mice; and (3) the response to sheep insulin is controlled by IA and IE encoded genes. These new findings differ from and extend previously published reports using bm12 mice, and therefore have substantive implications on the above stated conclusions regarding recognition of Ia. One such implication is that the bm12 gene conversion did not result in the transfer of a functional epitope for sheep insulin, but rather resulted in the creation of a functionally unique Ia molecule. Furthermore, this critical definition of the Ir gene lesion in bm12 permits us to address mechanistic questions regarding the nature of its Ir gene defect to beef insulin.